{"owner":"dusty-nv","repo":"jetson-containers","hasSkills":true,"totalSkillsCount":257,"totalTokensCount":548846,"categories":["anthropic-skill","root-instruction","claude-rule","plugin-manifest"],"hasMcp":false,"mcpConfig":null,"found":[".claude/skills/jetson-pr.md",".claude/skills/upgrade-package.md","AGENTS.md","CLAUDE.md","packages/README.md","packages/attention/ParaAttention/README.md","packages/attention/block-sparse-attention/README.md","packages/attention/flash-attention/README.md","packages/attention/flashinfer/README.md","packages/attention/flexprefill/README.md","packages/attention/huggingface_kernels/README.md","packages/attention/log-linear-attention/README.md","packages/attention/radial-attention/README.md","packages/attention/sage-attention/README.md","packages/attention/sparge-attention/README.md","packages/attention/xattention/README.md","packages/attention/xformers/README.md","packages/code/jupyterlab/README.md","packages/code/vscode/README.md","packages/cuda/cuda-python/README.md","packages/cuda/cuda/README.md","packages/cuda/cudastack/README.md","packages/cuda/cutlass/README.md","packages/cuda/pycuda/README.md","packages/cv/3d/3dobjects/partpacker/README.md","packages/cv/3d/3dobjects/sparc3d/README.md","packages/cv/3d/3dvision/glomap/README.md","packages/cv/3d/3dvision/hloc/README.md","packages/cv/3d/3dvision/kaolin/README.md","packages/cv/3d/3dvision/meshlab/README.md","packages/cv/3d/3dvision/nerfacc/README.md","packages/cv/3d/3dvision/open3d/README.md","packages/cv/3d/3dvision/pixsfm/README.md","packages/cv/3d/3dvision/polyscope/README.md","packages/cv/3d/3dvision/pyceres/README.md","packages/cv/3d/3dvision/pycolmap/README.md","packages/cv/3d/3dvision/pymeshlab/README.md","packages/cv/3d/3dvision/tinycudann/README.md","packages/cv/3d/gaussian_splatting/3dgrut/README.md","packages/cv/3d/gaussian_splatting/4k4d/README.md","packages/cv/3d/gaussian_splatting/easyvolcap/README.md","packages/cv/3d/gaussian_splatting/fast_gauss/README.md","packages/cv/3d/gaussian_splatting/gsplat/README.md","packages/cv/3d/gaussian_splatting/nerfview/README.md","packages/cv/3d/nerf/fruitnerf/README.md","packages/cv/3d/nerf/nerfstudio/README.md","packages/cv/cv-cuda/README.md","packages/cv/deepstream/README.md","packages/cv/diffusion/3d_diffusion_policy/README.md","packages/cv/diffusion/ai-toolkit/README.md","packages/cv/diffusion/cache_edit/README.md","packages/cv/diffusion/comfyui/README.md","packages/cv/diffusion/diffusers/README.md","packages/cv/diffusion/diffusion_policy/README.md","packages/cv/diffusion/fastgen/README.md","packages/cv/diffusion/framepack/README.md","packages/cv/diffusion/sdnext/README.md","packages/cv/diffusion/self-forcing/README.md","packages/cv/diffusion/stable-diffusion-webui/README.md","packages/cv/holoscan/README.md","packages/cv/jetson-inference/README.md","packages/cv/opencv/README.md","packages/cv/vpi/README.md","packages/hw/canable/README.md","packages/hw/canbus/README.md","packages/hw/cangaroo/README.md","packages/hw/jetcam/README.md","packages/hw/jupyter_clickable_image_widget/README.md","packages/hw/librealsense/README.md","packages/hw/oled/README.md","packages/hw/pl2303/README.md","packages/hw/zed/README.md","packages/llm/awq/README.md","packages/llm/bitsandbytes/README.md","packages/llm/deepspeed/README.md","packages/llm/deepspeed/deepspeed-kernels/README.md","packages/llm/dynamo/dynamo/README.md","packages/llm/dynamo/kai-scheduler/README.md","packages/llm/dynamo/mooncake/README.md","packages/llm/dynamo/nixl/README.md","packages/llm/exllama/README.md","packages/llm/gptqmodel/README.md","packages/llm/huggingface_hub/README.md","packages/llm/hymba/README.md","packages/llm/ktransformers/README.md","packages/llm/llama-factory/README.md","packages/llm/llama_cpp/README.md","packages/llm/llamaspeak/README.md","packages/llm/lobe_chat/README.md","packages/llm/local_llm/README.md","packages/llm/minference/README.md","packages/llm/mistral/mistral_rs/README.md","packages/llm/mlc/README.md","packages/llm/nano_llm/README.md","packages/llm/nemo/README.md","packages/llm/ollama/README.md","packages/llm/open-webui/README.md","packages/llm/openai/README.md","packages/llm/optimum/README.md","packages/llm/sglang/README.md","packages/llm/sglang/genai-bench/README.md","packages/llm/sglang/sgl-kernel/README.md","packages/llm/sudonim/README.md","packages/llm/tensorrt_optimizer/nvidia-modelopt/README.md","packages/llm/tensorrt_optimizer/tensorrt_llm/README.md","packages/llm/text-generation-inference/README.md","packages/llm/text-generation-webui/README.md","packages/llm/transformers/README.md","packages/llm/vllm/README.md","packages/llm/xgrammar/README.md","packages/ml/aim/README.md","packages/ml/apache/tvm/README.md","packages/ml/ctranslate2/README.md","packages/ml/dli/dli-nano-ai/README.md","packages/ml/jax/README.md","packages/ml/kans/kat/README.md","packages/ml/kans/pykan/README.md","packages/ml/l4t/l4t-diffusion/README.md","packages/ml/l4t/l4t-dynamo/README.md","packages/ml/l4t/l4t-ml/README.md","packages/ml/l4t/l4t-pytorch/README.md","packages/ml/l4t/l4t-tensorflow/README.md","packages/ml/l4t/l4t-text-generation/README.md","packages/ml/mamba/causalconv1d/README.md","packages/ml/mamba/cobra/README.md","packages/ml/mamba/dimba/README.md","packages/ml/mamba/mamba/README.md","packages/ml/mamba/mambavision/README.md","packages/ml/mamba/videomambasuite/README.md","packages/ml/mamba/zigma/README.md","packages/ml/numeric/arrow/README.md","packages/ml/numeric/cupy/README.md","packages/ml/numeric/numba/README.md","packages/ml/numeric/numpy/README.md","packages/ml/numeric/warp/README.md","packages/ml/onnx/README.md","packages/ml/onnxruntime/README.md","packages/ml/onnxruntime_genai/README.md","packages/ml/pytorch/README.md","packages/ml/pytorch/apex/README.md","packages/ml/pytorch/torch2trt/README.md","packages/ml/pytorch/torch3d/README.md","packages/ml/pytorch/torch_tensorrt/README.md","packages/ml/pytorch/torchao/README.md","packages/ml/pytorch/torchaudio/README.md","packages/ml/pytorch/torchsaver/README.md","packages/ml/pytorch/torchsde/README.md","packages/ml/pytorch/torchtext/README.md","packages/ml/pytorch/torchvision/README.md","packages/ml/rapids/cudf/README.md","packages/ml/rapids/cuml/README.md","packages/ml/rapids/raft/README.md","packages/ml/tensorflow/README.md","packages/ml/tensorflow/graphics/README.md","packages/ml/tensorflow/tensorboard/README.md","packages/ml/tensorflow/text/README.md","packages/ml/transformer-engine/README.md","packages/ml/triton/README.md","packages/ml/tritonserver/README.md","packages/ml/xlstm/mlstm_kernels/README.md","packages/ml/xlstm/pltsm/README.md","packages/ml/xlstm/xlstm/README.md","packages/multimedia/decord/README.md","packages/multimedia/ffmpeg/README.md","packages/multimedia/gstreamer/README.md","packages/multimedia/jetson-utils/README.md","packages/multimedia/libcom/README.md","packages/multimedia/opengl/README.md","packages/multimedia/pyav/README.md","packages/multimedia/sound-utils/README.md","packages/multimedia/video-codec-sdk/README.md","packages/multimedia/vulkan/README.md","packages/net/devpi/README.md","packages/net/pypi.dev/README.md","packages/physicalAI/cosmos/cosmos-policy/README.md","packages/physicalAI/cosmos/cosmos-predict2/README.md","packages/physicalAI/cosmos/cosmos-reason1/README.md","packages/physicalAI/cosmos/cosmos-transfer1/README.md","packages/physicalAI/cosmos/cosmos_diffusion_renderer/README.md","packages/physicalAI/cosmos/cosmos_diffusion_renderer/nvdiffrast/README.md","packages/physicalAI/isaac-ros/README.md","packages/physicalAI/lerobot/README.md","packages/physicalAI/opendronemap/README.md","packages/physicalAI/openpi/README.md","packages/physicalAI/protomotions/README.md","packages/physicalAI/ros/README.md","packages/physicalAI/sim/genesis/Readme.md","packages/physicalAI/sim/genesis/quadrants/README.md","packages/physicalAI/sim/genesis/splashSurf/README.md","packages/physicalAI/sim/genesis/vtk/README.md","packages/physicalAI/sim/habitat-sim/README.md","packages/physicalAI/sim/isaac-sim/README.md","packages/physicalAI/sim/isaac-sim/isaac-lab/README.md","packages/physicalAI/sim/mimicgen/README.md","packages/physicalAI/sim/mujoco/README.md","packages/physicalAI/sim/newton/README.md","packages/physicalAI/sim/physx/README.md","packages/physicalAI/sim/robogen/README.md","packages/physicalAI/sim/robomimic/README.md","packages/physicalAI/sim/robosuite/README.md","packages/physicalAI/sim/sapien/README.md","packages/physicalAI/vla/crossformer/README.md","packages/physicalAI/vla/isaac-gr00t/README.md","packages/physicalAI/vla/octo/README.md","packages/physicalAI/vla/openvla/README.md","packages/physicalAI/vla/robopoint/README.md","packages/rag/graphiti/README.md","packages/rag/jetson-copilot/README.md","packages/rag/langchain/README.md","packages/rag/llama-index/README.md","packages/rag/memvid/README.md","packages/smart-home/dependencies/ciso8601/README.md","packages/smart-home/dependencies/psutil-home-assistant/README.md","packages/smart-home/homeassistant-base/README.md","packages/smart-home/homeassistant-core/README.md","packages/smart-home/wyoming/README.md","packages/smart-home/wyoming/wyoming-assist-microphone/README.md","packages/smart-home/wyoming/wyoming-openwakeword/README.md","packages/smart-home/wyoming/wyoming-piper/README.md","packages/smart-home/wyoming/wyoming-whisper/README.md","packages/speech/audiocraft/README.md","packages/speech/chatterbox-tts/README.md","packages/speech/espeak/README.md","packages/speech/faster-whisper/README.md","packages/speech/kokoro-tts/kokoro-tts-fastapi/README.md","packages/speech/kokoro-tts/kokoro-tts-hf/README.md","packages/speech/kokoro-tts/kokoro-tts-onnx/README.md","packages/speech/piper1-tts/README.md","packages/speech/riva-client/README.md","packages/speech/spark-tts/README.md","packages/speech/speaches/README.md","packages/speech/speech-dispatcher/README.md","packages/speech/voice-pro/README.md","packages/speech/voicecraft/README.md","packages/speech/whisper/README.md","packages/speech/whisper_trt/README.md","packages/speech/whisperx/README.md","packages/speech/xtts/README.md","packages/vectordb/faiss/README.md","packages/vectordb/faiss_lite/README.md","packages/vectordb/nanodb/README.md","packages/vit/clip_trt/README.md","packages/vit/efficientvit/README.md","packages/vit/nanoowl/README.md","packages/vit/nanosam/README.md","packages/vit/sam/README.md","packages/vit/sapiens/README.md","packages/vit/tam/README.md","packages/vlm/lita/README.md","packages/vlm/llama-vision/README.md","packages/vlm/llava/README.md","packages/vlm/minigpt4/README.md","packages/vlm/prismatic/README.md","packages/vlm/videollama/README.md","packages/vlm/vila-microservice/README.md","packages/vlm/vila/README.md","packages/vlm/xtuner/README.md"],"skills":{".claude/skills/jetson-pr.md":"# Create a jetson-containers Pull Request\n\nHelp the user create a well-formed PR that follows this repository's conventions.\n\n## Repository PR conventions\n\n- **Target branch**: PRs go to `dev`, not `master`. The CI workflow `pr-on-dev.yml` runs self-hosted Jetson builds against `dev`.\n- **Scope**: each PR should touch only the packages it claims to change. Keep unrelated cleanup out.\n- **Commit style**: short, lowercase, imperative subject. No period. Examples from git log: `fix`, `cuda 13.2 needs newer drivers to run in Orin`, `tensorrt compatible orin`, `Update ffmpeg version from 8.1 to 8.0.1`.\n- **No force-push** to shared branches.\n\n## Step-by-step workflow\n\n1. **Understand the changes**: run `git diff master...HEAD` and `git log master..HEAD --oneline` to see exactly what is changing. Read the changed files if needed to understand the intent.\n\n2. **Identify affected packages**: map every changed file back to its package directory (`packages///`). List them — this drives the PR title and body.\n\n3. **Stage and commit** (if there are uncommitted changes):\n - Stage only the files related to this PR.\n - Write a commit message following the style above. No trailing period.\n - Do not skip hooks (`--no-verify`).\n\n4. **Create or switch to a feature branch** if on `master`:\n ```\n git checkout -b \n ```\n Branch name: lowercase, hyphenated, describes the change (e.g. `upgrade-pytorch-2.9`, `add-sglang-0.4`, `fix-tensorrt-orin`).\n\n5. **Push and open the PR**:\n ```\n gh pr create --base dev --title \"\" --body \"<body>\"\n ```\n\n6. **PR title format**: `<verb> <package(s)> <version or brief description>`\n - Version bumps: `upgrade pytorch to 2.9`\n - New packages: `add sglang 0.4`\n - Bug fixes: `fix tensorrt build on Orin`\n - Multi-package: list up to 3, then `and N others`\n\n7. **PR body template**:\n ```\n ## Changes\n - <package>: <what changed and why>\n\n ## Packages affected\n - `packages/<category>/<name>/`\n\n ## Test plan\n - [ ] `jetson-containers build <package>` on Orin (JP6)\n - [ ] Add any version-specific test steps here\n\n ## Notes\n <Any JetPack version constraints, driver requirements, or upstream release links>\n ```\n\n8. **Return the PR URL** when done.\n\n## What good PRs look like in this repo\n\n- Version upgrades include the new version entry in `config.py` **and** update `version.py` if it exists.\n- New packages include a `Dockerfile`, metadata (YAML header or `config.yaml`), and at minimum a stub `test.py` or `test.sh`.\n- Dependency bumps update `requires` constraints if the new version drops support for an older JetPack.\n- PRs do not mix version bumps with unrelated refactors.\n\n## Before opening the PR, verify\n\n- `git diff --stat dev...HEAD` shows only the intended files.\n- No `.env`, secrets, or build artifacts are staged.\n- The branch name and title are consistent.\n",".claude/skills/upgrade-package.md":"# Upgrade Package Version\n\nHelp the user upgrade the version of one or more packages in this jetson-containers repository.\n\n## What the user will provide\n\nThe user will tell you the package name and the new version. If they only give you the package name, ask for the target version before proceeding.\n\n## How version definitions work\n\nEach package under `packages/<category>/<name>/` defines its version in one of these ways:\n\n1. **`version.py`** — a standalone file that exports a version constant, e.g. `PYTORCH_VERSION = Version('2.8')`. This is the canonical version used by `config.py`.\n2. **`config.py`** — calls a helper function (like `pytorch_pip(version, requires=...)`) with explicit version strings in a `package = [...]` list. Adding a new version = adding a new line to this list.\n3. **`config.yaml` / YAML header in Dockerfile** — static `build_args` with version pinned directly.\n4. **`Dockerfile` `ARG` lines** — `ARG SOME_VERSION=x.y.z`.\n\n## Step-by-step workflow\n\n1. **Locate the package directory**: search under `packages/` for the package name.\n2. **Read all config files** in that directory (`config.py`, `config.yaml`, `version.py`, `Dockerfile`, `Dockerfile.*`).\n3. **Identify where the version is defined** (see above). Show the user the current version(s) before making changes.\n4. **Apply the upgrade**:\n - If there is a `version.py`, update the version constant there first; `config.py` will pick it up automatically.\n - If `config.py` uses a list of `pytorch_pip(version, ...)` calls, **add** a new entry for the new version (do not remove old ones — the repo keeps historical versions for different JetPack targets). Set the new version as the latest by updating the `PYTORCH_VERSION`-like constant in `version.py`.\n - If the version is in a YAML `build_args` or a Dockerfile `ARG`, do a targeted edit of that value.\n5. **Check for cascading dependencies**: grep for the old version string across other config files in the same package directory. Common locations: `Dockerfile` `ARG` defaults, `build_args` dicts, `requires` strings, download URLs, wheel filenames.\n6. **Update download artifacts if needed**: if the Dockerfile fetches a wheel, tarball, or pre-built binary, update the URL/filename to match the new version. Check the upstream source (PyPI, NVIDIA, GitHub releases) if the user hasn't provided a URL.\n7. **Show a diff summary** of every file you changed before finishing.\n\n## Things to preserve\n\n- Do **not** delete existing version entries in `config.py` lists — old JetPack targets still need them.\n- Keep the `requires` constraints on old entries unchanged.\n- Do not change `alias` lists for older versions.\n- Follow the existing code style (no reformatting unrelated code, no added comments).\n\n## When to ask vs. proceed\n\n- Ask for the target version if not given.\n- Ask about JetPack `requires` constraint only if the new version introduces a minimum JetPack requirement that differs from the previous latest entry.\n- Proceed with everything else silently and show the diff at the end.\n","AGENTS.md":"# CLAUDE.md\n\nThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.\n\n## Overview\n\njetson-containers is a modular container build system for NVIDIA Jetson devices. It discovers packages, resolves their dependencies, chains their Dockerfiles together, and runs `docker build` to produce multi-stage images optimized for specific L4T/JetPack/CUDA versions.\n\n## Commands\n\n### Build & Run\n```bash\n# Build one or more packages (chains Dockerfiles in dependency order)\njetson-containers build <package> [<package2> ...]\n# or equivalently:\n./build.sh <package> [<package2> ...]\n\n# Find and run a compatible container image\njetson-containers run $(autotag <package>) [command]\n# or equivalently:\n./run.sh $(autotag <package>) [command]\n\n# Find the best matching image tag (local → DockerHub → build)\nautotag <package>\n\n# Useful build flags\njetson-containers build --simulate <package> # dry-run: print commands without building\njetson-containers build --multiple pytorch tensorflow # build each package as a separate image\njetson-containers build --list-packages # list all discovered packages\njetson-containers build --show-packages <package> # show resolved metadata for a package\njetson-containers build --base=xyz:latest <package> # start the build chain from an existing image\njetson-containers build --push=myuser <package> # push to a registry after building\n\n# Check for upstream version upgrades (--apply to write changes)\njetson-containers upgrade [<package> ...] [--apply]\n```\n\n### Code Style\n```bash\n# Install pre-commit hooks (one-time setup)\npip install -r requirements.txt\npre-commit install\n\n# Run linting/formatting checks manually\npre-commit run --all-files\n```\n\nPre-commit runs **Black** (formatter, line-length=88) and **Flake8** (linter) only on `jetson_containers/` and `test_precommit.py`. `packages/` Dockerfiles and `config.py` files are **not** linted by pre-commit. Note: `skip-string-normalization = true` in `pyproject.toml`.\n\n### Tests\n```bash\n# Run a package config unit test directly (uses stdlib unittest, no pytest needed)\npython3 tests/verify_triton_config.py\n```\n\n### Installation\n```bash\nbash install.sh # sets up venv, installs deps, symlinks CLI tools to /usr/local/bin\n```\n\n## Architecture\n\n### Package System\n\nEvery buildable unit is a **package**. Packages live under `packages/<category>/<name>/` and define metadata via one of (in priority order):\n\n1. **YAML header in Dockerfile** — metadata between `#---` markers at the top\n2. **`config.yaml` / `config.yml`** — static YAML config file\n3. **`config.json`** — meta-container (no Dockerfile, just dependency composition)\n4. **`config.py`** — dynamic config, returns a list of package dicts at build time\n\nKey metadata fields: `name`, `alias`, `depends` (build-order deps), `requires` (version constraints on L4T/CUDA/Python), `test`, `build_args`, `dockerfile`.\n\n### Build Pipeline\n\n1. **`jetson_containers/packages.py`** — scans `packages/` recursively, runs `config.py` scripts, resolves the dependency graph, filters by L4T version compatibility, and exposes a custom Python meta path finder so packages can `from packages.ml.pytorch.version import ...`\n2. **`jetson_containers/build.py`** — CLI entry point; parses args and dispatches\n3. **`jetson_containers/container.py`** — core build logic: stitches multiple package Dockerfiles into one combined `Dockerfile`, invokes `docker build`, optionally runs per-package tests, and optionally pushes to a registry\n4. **`jetson_containers/l4t_version.py`** — detects architecture (tegra-aarch64, aarch64, x86_64) and reads L4T/JetPack/CUDA/GPU-arch info from `/etc/nv_tegra_release`; all version variables (e.g. `CUDA_VERSION`, `PYTORCH_VERSION`) flow from here into build args\n\n### Dynamic Package Generation\n\n`config.py` files execute at build time and can return multiple package dicts to produce version variants. For example, PyTorch's `config.py` generates `pytorch:2.8`, `pytorch:2.8-all`, `pytorch:2.8-builder`, etc. Aliases (`torch` → `pytorch:2.8`) are resolved during package discovery.\n\nWhen a package's version selection logic is complex, put it in a separate `version.py` alongside `config.py`. This lets other packages cross-import it:\n```python\nfrom packages.ml.pytorch.version import PYTORCH_VERSION\n```\nThe `_PackagesFinder` meta path finder in `packages.py` resolves these imports against the `packages/` tree without requiring `__init__.py` files.\n\n### Dockerfile Conventions\n\nUse `uv pip install` (not plain `pip install`) inside Dockerfiles — `uv` is pre-installed in the base images and is significantly faster. Every Dockerfile must start with:\n```dockerfile\nARG BASE_IMAGE\nFROM ${BASE_IMAGE}\n```\n\n### Version Overrides\n\nCopy `.env.default` → `.env` to override `L4T_VERSION`, `CUDA_VERSION`, `LSB_RELEASE`, `CUDA_ARCH`, local PyPI/APT mirror URLs, and SCP/webhook credentials without modifying tracked files.\n\n### Creating a New Package\n\n1. Create `packages/<category>/<name>/Dockerfile` with a `#---` YAML header (or a `config.yaml`/`config.py`).\n2. Declare `depends` on upstream packages (e.g., `pytorch`, `cuda`).\n3. Optionally add `test.py` or `test.sh` for post-build validation.\n4. Build: `jetson-containers build <name>`\n\n## Security Features\n\n### Multi-Stage Builds\n\nThe build system is multi-stage aware. Packages can use `FROM ... AS <alias>` for intermediate stages (e.g. a compiler/builder stage) without those stages polluting the final image. The build system tracks stage boundaries and only injects BuildKit features (ccache env vars, `--device` mounts, ccache `--mount` flags) into the **final** stage (the `FROM ${BASE_IMAGE}` stage). Intermediate stages are left untouched.\n\nDeclare multi-stage packages in config metadata:\n```yaml\n#---\n# name: mypackage\n# multistage: true\n# depends: [cuda]\n#---\nARG BASE_IMAGE\nFROM gcc:12 AS builder\nRUN make install\n\nFROM ${BASE_IMAGE}\nCOPY --from=builder /usr/local /usr/local\nRUN ldconfig\n```\n\nThe final stage **must** use `ARG BASE_IMAGE` / `FROM ${BASE_IMAGE}` so it chains correctly in the dependency build pipeline.\n\n### Non-Root User\n\n**At build time** — pass identity via env vars; packages can create and switch to the declared user:\n```bash\nexport CONTAINER_USER=jetson\nexport CONTAINER_UID=1000\nexport CONTAINER_GID=1000\njetson-containers build mypackage\n```\n\n`CONTAINER_USER`, `CONTAINER_UID`, `CONTAINER_GID` are forwarded as `--build-arg` automatically. Packages consume them like:\n```dockerfile\nARG CONTAINER_USER=root\nARG CONTAINER_UID=0\nARG CONTAINER_GID=0\nRUN if [ \"$CONTAINER_UID\" != \"0\" ]; then \\\n groupadd -g $CONTAINER_GID $CONTAINER_USER && \\\n useradd -u $CONTAINER_UID -g $CONTAINER_GID -m $CONTAINER_USER; \\\n fi\nUSER $CONTAINER_USER\n```\n\nDeclare the supported user in package metadata (`run_user` is informational for tooling):\n```yaml\n# run_user: jetson\n```\n\n**At runtime** — set `DOCKER_USER` to run the container as a non-root identity:\n```bash\nDOCKER_USER=1000:1000 ./run.sh myimage\n# or by name (the user must exist in the image):\nDOCKER_USER=jetson ./run.sh myimage\n```\nX11 forwarding (`xhost`) is automatically updated to grant the declared user display access.\n\nFor GPU access, `run.sh` also discovers the numeric GIDs that own the NVIDIA/Tegra device nodes (`/dev/nvhost*`, `/dev/nvmap`, `/dev/nvgpu/*`, `/dev/dri/render*`) and adds them via `--group-add <gid>`. Without this, a non-root process can't open the GPU nodes and CUDA fails with error 801 (`cudaGetDeviceCount` \"operation not supported\"). Numeric GIDs are used because the host's `render` GID usually doesn't match the container's. See [`docs/run.md`](/docs/run.md#gpu-access-as-non-root-cuda-error-801).\n\n### Secrets\n\n**Build-time secrets** (never baked into image layers) — declare secrets in package metadata:\n```yaml\n#---\n# name: mypackage\n# secrets: [huggingface_token]\n#---\n```\n\nPoint each secret at a file on the host via `JETSON_SECRET_<NAME>`:\n```bash\nexport JETSON_SECRET_HUGGINGFACE_TOKEN=~/.hf_token\njetson-containers build mypackage\n```\n\nSecrets are mounted as `/run/secrets/<name>` inside `RUN` steps only (BuildKit `--secret`). Use them in the Dockerfile:\n```dockerfile\nRUN --mount=type=secret,id=huggingface_token \\\n HF_TOKEN=$(cat /run/secrets/huggingface_token) \\\n python3 download_model.py\n```\n\nIf the env var is missing, a warning is printed and the build continues without that secret.\n\n**Runtime secrets** — prefer file-backed tokens over env vars (values in env vars are visible in `docker inspect`):\n```bash\n# Preferred: mount token file into /run/secrets/\nHUGGINGFACE_TOKEN_FILE=~/.hf_token ./run.sh myimage\n\n# Fallback (token visible in docker inspect):\nHUGGINGFACE_TOKEN=hf_xxx ./run.sh myimage\n```\n\n`HUGGINGFACE_TOKEN_FILE` mounts the file read-only at `/run/secrets/huggingface_token` and sets `HF_TOKEN_FILE` inside the container.\n\n### Key Supporting Modules\n\n- `jetson_containers/tag.py` — image tagging conventions\n- `jetson_containers/docs.py` — auto-generates per-package docs from metadata\n- `jetson_containers/ci.py` — CI/CD helpers\n- `jetson_containers/network.py` / `webhook.py` — GitHub API integration and build notifications\n- `jetson_containers/logging.py` — colored terminal output used throughout\n- `jetson_containers/upgrade.py` — checks packages against upstream (PyPI/GitHub) for newer versions; powers `jetson-containers upgrade`\n- `jetson_containers/db.py` — syncs DockerHub/GitHub metadata for registry introspection\n","CLAUDE.md":"@AGENTS.md\n","packages/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FREADME.md -->","packages/attention/ParaAttention/README.md":"# ParaAttention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nhttps://github.com/chengzeyi/ParaAttention\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`paraattention:0.4.0`** | |\n| :-- | :-- |\n|    Aliases | `paraattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ParaAttention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ParaAttention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ParaAttention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ParaAttention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ParaAttention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/block-sparse-attention/README.md":"# block-sparse-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`block_sparse_attn:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `block_sparse_attn` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`xattention:0.0.1`](/packages/attention/xattention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag block-sparse-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host block-sparse-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag block-sparse-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag block-sparse-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build block-sparse-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flash-attention/README.md":"# flash-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flash-attention:2.5.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.6.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.4.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.0.post2`** | |\n| :-- | :-- |\n|    Aliases | `flash-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flashinfer/README.md":"# flash-infer\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flashinfer:0.2.1.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:0.8.4`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.6.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`mlc:0.20.0`](/packages/llm/mlc) [`sglang:0.4.6`](/packages/llm/sglang) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.8`** | |\n| :-- | :-- |\n|    Aliases | `flashinfer` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`llama-factory`](/packages/llm/llama-factory) [`local_llm`](/packages/llm/local_llm) [`mlc:0.21.0`](/packages/llm/mlc) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-infer)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-infer:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-infer)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-infer) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-infer\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flexprefill/README.md":"# flexprefill\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flexprefill:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `flexprefill` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`llama-factory`](/packages/llm/llama-factory) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`sdnext`](/packages/diffusion/sdnext) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flexprefill)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flexprefill:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flexprefill)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flexprefill) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flexprefill\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/huggingface_kernels/README.md":"# huggingface_kernels\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`huggingface_kernels:0.7.0`** | |\n| :-- | :-- |\n|    Aliases | `huggingface_kernels` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`huggingface_hub`](/packages/llm/huggingface_hub) [`torchvision`](/packages/pytorch/torchvision) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag huggingface_kernels)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host huggingface_kernels:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag huggingface_kernels)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag huggingface_kernels) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build huggingface_kernels\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/log-linear-attention/README.md":"# log-linear-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`log-linear-attention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `log-linear-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`ninja`](/packages/build/ninja) [`torchaudio`](/packages/pytorch/torchaudio) [`causalconv1d`](/packages/ml/mamba/causalconv1d) [`mamba`](/packages/ml/mamba/mamba) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag log-linear-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host log-linear-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag log-linear-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag log-linear-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build log-linear-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/radial-attention/README.md":"# radial-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`radial-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `radial-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`flashinfer`](/packages/attention/flash-infer) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag radial-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host radial-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag radial-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag radial-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build radial-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/sage-attention/README.md":"# sage-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sage-attention:3.0.0`** | |\n| :-- | :-- |\n|    Aliases | `sage-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`framepack`](/packages/diffusion/framepack) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sage-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sage-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sage-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sage-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sage-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/sparge-attention/README.md":"# sparge-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparge-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparge-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/thu-ml/SpargeAttn.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparge-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparge-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparge-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparge-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparge-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/xattention/README.md":"# xattention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xattention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `xattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`block_sparse_attn`](/packages/attention/block-sparse-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xattention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host xattention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xattention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xattention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xattention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/xformers/README.md":"# xformers\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xformers:0.0.32`** | |\n| :-- | :-- |\n|    Aliases | `xformers` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dependants | [`apex:0.1`](/packages/pytorch/apex) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n| **`xformers:0.0.33`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/xformers:r35.2.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-06` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.3.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-14` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.4.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2024-01-09` | `arm64` | `5.9GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xformers)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/xformers:r35.4.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/xformers:r35.4.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xformers)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xformers) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xformers\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/code/jupyterlab/README.md":"# jupyterlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nRunning the `jupyerlab` container will automatically start a JupyterLab server in the background on port 8888, with default login password `nvidia`. The JupyterLab server logs will be saved to `/data/logs/jupyter.log` if you need to inspect them (this location is automatically mounted under your `jetson-containers/data` directory)\n\nTo change the default settings, you can set the `$JUPYTER_ROOT`, `$JUPYTER_PORT`, `$JUPYTER_PASSWORD`, and `$JUPYTER_LOG` environment variables when starting the container like so:\n\n```bash\njetson-containers run \\\n --env JUPYTER_ROOT=/home/user \\\n --env JUPYTER_PORT=8000 \\\n --env JUPYTER_PASSWORD=password \\\n --env JUPYTER_LOGS=/dev/null \\\n $(autotag jupyterlab)\n```\n\nThe [`/start_jupyter`](./start_jupyter) script is the default CMD that the container runs when it starts - however, if you don't want the JupyterLab server started by default, you can either add a different CMD in your own Dockerfile, or override it at startup:\n\n```bash\n# skip straight to the terminal instead of starting JupyterLab first\njetson-containers run /bin/bash\n```\n\nYou can then still manually run the [`/start_jupyter`](./start_jupyter) script later when desired.\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`jupyterlab:latest`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:latest-myst`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab:myst` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:latest`](/packages/code/jupyterlab) |\n|    Dependants | [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0-myst`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/jupyterlab:r32.7.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.7GB` |\n|   [`dustynv/jupyterlab:r35.2.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-12-06` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r35.3.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `5.4GB` |\n|   [`dustynv/jupyterlab:r35.4.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-10-07` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r36.2.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.6GB` |\n|   [`dustynv/jupyterlab:r36.4.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-10` | `arm64` | `5.8GB` |\n|   [`dustynv/jupyterlab:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-03` | `arm64` | `5.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag jupyterlab)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/jupyterlab:r36.4.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/jupyterlab:r36.4.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag jupyterlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag jupyterlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build jupyterlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/code/vscode/README.md":"# vscode\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`vscode`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:cuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`tensorrt`](/packages/cuda/tensorrt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:torch`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`tensorrt`](/packages/cuda/tensorrt) [`torch2trt`](/packages/pytorch/torch2trt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:transformers`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`bitsandbytes`](/packages/llm/bitsandbytes) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag vscode)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host vscode:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag vscode)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag vscode) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build vscode\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/cuda-python/README.md":"# cuda-python\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda-python:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda-python` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cuml`](/packages/ml/rapids/cuml) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss_lite`](/packages/vectordb/faiss_lite) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index:samples`](/packages/rag/llama-index) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`raft`](/packages/ml/rapids/raft) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`tam`](/packages/vit/tam) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.8`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu128`](/packages/cuda/cuda) [`cuda:12.8`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.9`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu129`](/packages/cuda/cuda) [`cuda:12.9`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:13.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu130`](/packages/cuda/cuda) [`cuda:13.0`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/cuda-python:builder-r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-26` | `arm64` | `5.1GB` |\n|   [`dustynv/cuda-python:builder-r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-11` | `arm64` | `3.6GB` |\n|   [`dustynv/cuda-python:r35.2.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.3.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/cuda-python:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cuda-python)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/cuda-python:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/cuda-python:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cuda-python)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cuda-python) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cuda-python\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/cuda/README.md":"# cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/cuda:12.2-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-05, 3.4GB)`<br>[`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n\n| **`pip_cache:cu122`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2`](/packages/cuda/cuda) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.2-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Images | [`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu124`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4`](/packages/cuda/cuda) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.4-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda` |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`clip_trt`](/packages/vit/clip_trt) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pyav`](/packages/multimedia/pyav) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/packages/speech/whisperx) [`wyoming-piper:1.6.2`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-piper:master`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-whisper:2.5.0`](/packages/smart-home/wyoming/wyoming-whisper) [`wyoming-whisper:master`](/packages/smart-home/wyoming/wyoming-whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xformers:0.0.32`](/packages/attention/xformers) [`xformers:0.0.33`](/packages/attention/xformers) [`xgrammar:0.1.15`](/packages/llm/xgrammar) [`xgrammar:0.1.18`](/packages/llm/xgrammar) [`xgrammar:0.1.19`](/packages/llm/xgrammar) [`xgrammar:0.1.20`](/packages/llm/xgrammar) [`xgrammar:0.1.21`](/packages/llm/xgrammar) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtts`](/packages/speech/xtts) [`xtuner`](/packages/vlm/xtuner) [`zed:5.0`](/packages/hw/zed) [`zed:5.0-humble`](/packages/hw/zed) [`zed:5.0-jazzy`](/packages/hw/zed) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu126`** | |\n| :-- | :-- |\n|    Aliases | `pip_cache` `pip-cache` |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`aim`](/packages/ml/aim) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`canbus`](/packages/hw/canbus) [`cangaroo`](/packages/hw/cangaroo) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`ciso8601:v2.3.2`](/packages/smart-home/dependencies/ciso8601) [`clip_trt`](/packages/vit/clip_trt) [`cmake:pip`](/packages/build/cmake/cmake_pip) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6`](/packages/cuda/cuda) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`docker`](/packages/build/docker) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`graphiti`](/packages/rag/graphiti) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`h5py`](/packages/build/h5py) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`homeassistant-core:2025.7.0`](/packages/smart-home/homeassistant-core) [`huggingface_hub`](/packages/llm/huggingface_hub) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:python`](/packages/multimedia/jetson-utils) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llamaspeak`](/packages/llm/llamaspeak) [`llava`](/packages/vlm/llava) [`llvm:10`](/packages/build/llvm) [`llvm:11`](/packages/build/llvm) [`llvm:12`](/packages/build/llvm) [`llvm:13`](/packages/build/llvm) [`llvm:14`](/packages/build/llvm) [`llvm:15`](/packages/build/llvm) [`llvm:16`](/packages/build/llvm) [`llvm:17`](/packages/build/llvm) [`llvm:18`](/packages/build/llvm) [`llvm:19`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`meshlab:MeshLab-2023.12`](/packages/3d/3dvision/meshlab) [`meshlab:MeshLab-2025.03`](/packages/3d/3dvision/meshlab) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`mujoco`](/packages/sim/mujoco) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`ninja`](/packages/build/ninja) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`numpy`](/packages/numeric/numpy) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnx`](/packages/ml/onnx) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`openai`](/packages/llm/openai) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`polyscope:2.5.0`](/packages/3d/3dvision/polyscope) [`prismatic`](/packages/vlm/prismatic) [`protobuf:apt`](/packages/build/protobuf/protobuf_apt) [`protobuf:cpp`](/packages/build/protobuf/protobuf_cpp) [`protomotions:2.5.0`](/packages/robots/protomotions) [`psutil-home-assistant:0.0.1`](/packages/smart-home/dependencies/psutil-home-assistant) [`pyav`](/packages/multimedia/pyav) [`pybind11`](/packages/build/pybind11) [`pyceres:2.5`](/packages/3d/3dvision/pyceres) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`python:3.10`](/packages/build/python) [`python:3.11`](/packages/build/python) [`python:3.12`](/packages/build/python) [`python:3.13`](/packages/build/python) [`python:3.14`](/packages/build/python) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`riva-client:python`](/packages/speech/riva-client) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`rust`](/packages/build/rust) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim`](/packages/llm/sudonim) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorboard`](/packages/ml/tensorflow/tensorboard) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode`](/packages/code/vscode) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`vtk:9.3.1`](/packages/sim/genesis/vtk) [`vtk:9.4.2`](/packages/sim/genesis/vtk) [`vtk:9.5.0`](/packages/sim/genesis/vtk) [`vulkan:1.3.275`](/packages/multimedia/vulkan) [`vulkan:1.3.296`](/packages/multimedia/vulkan) [`vulkan:1.4.309`](/packages/multimedia/vulkan) [`vulkan:1.4.313`](/packages/multimedia/vulkan) [`vulkan:1.4.314`](/packages/multimedia/vulkan) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/pa\n\n<!-- [TRUNCATED AT 50KB] -->","packages/cuda/cudastack/README.md":"# CUDA Stack - Consolidated Library Installation\n\n## What is cudastack?\n\n`cudastack` is a **consolidation package** that installs multiple CUDA libraries (cuDNN, NCCL, TensorRT, CUTLASS, etc.) in **ONE RUN command** to avoid Docker's \"max depth exceeded\" error.\n\n## Important: cudastack vs cuda\n\n```\ncuda/ cudastack/\n├── Dockerfile ├── Dockerfile (ONE RUN for all libs)\n├── Dockerfile.pip ├── config.py (variants)\n├── Dockerfile.builtin └── install/ & build/ scripts\n├── Dockerfile.samples\n└── config.py\n ├── cuda_package() cudastack DEPENDS ON cuda\n ├── cuda_builtin() (doesn't replace it!)\n ├── cuda_samples()\n └── pip_cache()\n```\n\n**Key Points:**\n\n1. ✅ **`cuda` package** - Installs CUDA toolkit, provides pip_cache, samples, builtin variants\n2. ✅ **`cudastack` package** - Builds ON TOP of `cuda`, adds cuDNN/NCCL/TensorRT/etc. in ONE layer\n3. ✅ **pip_cache, cuda_samples, cuda_builtin** - Still work! They're part of `cuda`, not `cudastack`\n\n## The Problem It Solves\n\n### Before (Separate Packages):\n```\nLayer 1: cuda\nLayer 2: cudnn\nLayer 3: tensorrt\nLayer 4: nccl\nLayer 5: cutlass\n...\nLayer 50+: \"max depth exceeded\" ❌\n```\n\n### After (cudastack):\n```\nLayer 1: cuda (with pip_cache, samples, etc.)\nLayer 2: cudastack (cudnn + nccl + tensorrt + cutlass in ONE RUN)\nLayer 3: your-app\nResult: Works! ✅\n```\n\n## Architecture\n\n```\nBASE_IMAGE\n ↓\ncuda:12.4 (from cuda/Dockerfile)\n ├── pip_cache:cu124 (from cuda/Dockerfile.pip)\n └── cuda:12.4-samples (from cuda/Dockerfile.samples)\n ↓\ncudastack:minimal (from cudastack/Dockerfile)\n ├── cuDNN\n ├── NCCL\n └── (all installed in ONE RUN)\n ↓\nyour-app\n```\n\n## Variants\n\n| Variant | Includes | Use Case |\n|---------|----------|----------|\n| `cudastack:minimal` | cuDNN + NCCL | Training, inference |\n| `cudastack:standard` | + TensorRT | Optimized inference |\n| `cudastack:full` | + CUTLASS + GDRCopy | Research, development |\n\n## Usage\n\n### Build with jetson-containers:\n\n```bash\n# Build minimal stack\n./build.sh cudastack:minimal\n\n# Build standard (with TensorRT)\n./build.sh cudastack:standard\n\n# Build full (with CUTLASS)\n./build.sh cudastack:full\n\n# Use as base for your app\n./build.sh --base cudastack:minimal pytorch\n```\n\n### The cudastack Dockerfile does ONE RUN:\n\n```dockerfile\n# Copy all install/build scripts\nCOPY install/*.sh /tmp/cudastack/install/\nCOPY build/*.sh /tmp/cudastack/build/\n\n# ONE RUN command calls all scripts conditionally\nRUN set -ex && \\\n if [ \"$WITH_CUDNN\" = \"1\" ]; then \\\n ./install/install_cudnn.sh; \\\n fi && \\\n if [ \"$WITH_NCCL\" = \"1\" ]; then \\\n ./install/install_nccl.sh; \\\n fi && \\\n if [ \"$WITH_TENSORRT\" = \"1\" ]; then \\\n ./install/install_tensorrt.sh; \\\n fi && \\\n # ... more components ...\n rm -rf /tmp/cudastack\n```\n\n**Key**: All library installations happen in this ONE RUN = ONE Docker layer!\n\n## Setup\n\nThe install scripts need to be populated from existing packages:\n\n```bash\ncd packages/cuda/cudastack\n./setup-scripts.sh\n```\n\nThen review and complete the scripts in `install/` and `build/`.\n\n## What About pip_cache, samples, builtin?\n\nThese are **still available** from the `cuda` package:\n\n```bash\n# pip_cache is automatically included by cuda package\n./build.sh cuda:12.4\n# This creates both:\n# - cuda:12.4 (main CUDA toolkit)\n# - pip_cache:cu124 (pip caching)\n\n# CUDA samples work as before\n./build.sh cuda:12.4-samples\n\n# Builtin works as before (for JetPack 4-5)\n./build.sh cuda:11.4 # Uses Dockerfile.builtin when CUDA pre-installed\n```\n\nThe `cuda` package's `config.py` returns multiple packages:\n```python\n# In cuda/config.py\ndef cuda_package(version, url, ...):\n cuda = package.copy()\n cuda['name'] = f'cuda:{version}'\n \n cuda_pip = pip_cache(version, requires) # Creates pip_cache package\n \n return cuda, cuda_pip # Returns TWO packages!\n```\n\n**cudastack doesn't need to duplicate this** - it just depends on `cuda` being installed first.\n\n## Component Toggles\n\nControlled by `WITH_*` build args (set in config.py):\n\n- ✅ `WITH_CUDNN=1` - cuDNN (always on)\n- ✅ `WITH_NCCL=1` - NCCL (always on)\n- 🎛️ `WITH_TENSORRT` - TensorRT (on in standard/full)\n- 🎛️ `WITH_CUTLASS` - CUTLASS (on in full only)\n- 🎛️ `WITH_GDRCOPY` - GDRCopy (on for Tegra)\n- ⚙️ `WITH_CUDSS=0` - cuDSS (off by default)\n- ⚙️ `WITH_CUSPARSELT=0` - cuSPARSELt (off by default)\n- ⚙️ `WITH_CUTENSOR=0` - cuTENSOR (off by default)\n- ⚙️ `WITH_NVPL=0` - NVPL (off by default)\n- ⚙️ `WITH_NVSHMEM=0` - NVSHMEM (off by default)\n\n## How config.py Works\n\nFollowing the jetson-containers pattern:\n\n```python\ndef cuda_stack(name, with_tensorrt=False, with_cutlass=False, requires=None):\n \"\"\"Generate a cudastack package variant\"\"\"\n pkg = package.copy()\n pkg['name'] = name\n pkg['depends'] = ['cuda'] # Depends ON cuda, doesn't replace it\n \n pkg['build_args'] = {\n 'WITH_CUDNN': '1',\n 'WITH_TENSORRT': '1' if with_tensorrt else '0',\n 'WITH_NCCL': '1',\n # ... etc\n }\n \n return pkg\n\n# Define variants\npackage = [\n cuda_stack('cudastack:minimal', with_tensorrt=False),\n cuda_stack('cudastack:standard', with_tensorrt=True),\n cuda_stack('cudastack:full', with_tensorrt=True, with_cutlass=True),\n]\n```\n\n## Comparison with Individual Packages\n\n| Approach | Layers | Build Time | Complexity |\n|----------|--------|------------|------------|\n| **Individual packages** | 50+ | Longer | Simple per-package |\n| **cudastack (consolidated)** | 2-3 | Faster | One complex layer |\n\nBoth approaches work! Use `cudastack` when you need multiple libraries and want to avoid layer limits.\n\n## See Also\n\n- [cuda package](../cuda/) - CUDA toolkit, pip_cache, samples, builtin\n- [cudnn package](../cudnn/) - Individual cuDNN package\n- [tensorrt package](../tensorrt/) - Individual TensorRT package (has Dockerfile.deb and Dockerfile.tar)\n- [nccl package](../nccl/) - Individual NCCL package\n\n## Summary\n\n✅ **cudastack is a consolidation layer** on top of cuda \n✅ **pip_cache, samples, builtin still work** from the cuda package \n✅ **ONE RUN command** installs all libraries to avoid layer limits \n✅ **Follows jetson-containers patterns** with config.py and variants \n","packages/cuda/cutlass/README.md":"# cutlass\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cutlass:4.1.0`** | |\n| :-- | :-- |\n|    Aliases | `cutlass` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cutlass)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cutlass:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cutlass)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cutlass) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cutlass\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/pycuda/README.md":"# pycuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 0.4GB)`<br>[`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-09-07, 5.0GB)`<br>[`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-08-29, 5.0GB)`<br>[`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 5.0GB)`<br>[`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 3.5GB)`<br>[`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2025-03-03, 2.3GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `0.4GB` |\n|   [`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-09-07` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycuda)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/pycuda:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/pycuda:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dobjects/partpacker/README.md":"# partpacker\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`partpacker:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `partpacker` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/PartPacker |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag partpacker)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host partpacker:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag partpacker)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag partpacker) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build partpacker\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dobjects/sparc3d/README.md":"# sparc3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparc3d:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparc3d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/lizhihao6/Sparc3D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparc3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparc3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparc3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparc3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparc3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/glomap/README.md":"# glomap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`glomap:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `glomap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag glomap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host glomap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag glomap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag glomap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build glomap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/hloc/README.md":"# hloc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`hloc:1.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n| **`hloc:1.5`** | |\n| :-- | :-- |\n|    Aliases | `hloc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag hloc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host hloc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag hloc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag hloc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build hloc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/kaolin/README.md":"# kaolin\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`kaolin:0.18.0`** | |\n| :-- | :-- |\n|    Aliases | `kaolin` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVIDIAGameWorks/kaolin |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag kaolin)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host kaolin:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag kaolin)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag kaolin) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build kaolin\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/meshlab/README.md":"# meshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`meshlab:MeshLab-2023.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n| **`meshlab:MeshLab-2025.03`** | |\n| :-- | :-- |\n|    Aliases | `meshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag meshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host meshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag meshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag meshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build meshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/nerfacc/README.md":"# nerfacc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfacc:0.5.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n| **`nerfacc:0.5.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfacc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfacc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfacc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfacc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfacc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfacc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/open3d/README.md":"# open3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`open3d:1.19.0`** | |\n| :-- | :-- |\n|    Aliases | `open3d` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag open3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host open3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag open3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag open3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build open3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pixsfm/README.md":"# pixsfm\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pixsfm:1.0`** | |\n| :-- | :-- |\n|    Aliases | `pixsfm` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pixsfm)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pixsfm:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pixsfm)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pixsfm) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pixsfm\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/polyscope/README.md":"# polyscope\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`polyscope:2.5.0`** | |\n| :-- | :-- |\n|    Aliases | `polyscope` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nmwsharp/polyscope-py.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag polyscope)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host polyscope:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag polyscope)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag polyscope) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build polyscope\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pyceres/README.md":"# pyceres\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pyceres:2.5`** | |\n| :-- | :-- |\n|    Aliases | `pyceres` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`ninja`](/packages/build/ninja) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/pyceres.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pyceres)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pyceres:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pyceres)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pyceres) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pyceres\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pycolmap/README.md":"# pycolmap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycolmap:3.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n| **`pycolmap:3.13`** | |\n| :-- | :-- |\n|    Aliases | `pycolmap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycolmap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pycolmap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycolmap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycolmap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycolmap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pymeshlab/README.md":"# pymeshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pymeshlab:2023.12.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2023.12.post3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2025.6.23.dev0`** | |\n| :-- | :-- |\n|    Aliases | `pymeshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pymeshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pymeshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pymeshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pymeshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pymeshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/tinycudann/README.md":"# tinycudann\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`tinycudann:1.7`** | |\n| :-- | :-- |\n|    Aliases | `tinycudann` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/tiny-cuda-nn |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag tinycudann)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host tinycudann:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag tinycudann)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag tinycudann) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build tinycudann\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/3dgrut/README.md":"# 3dgrut\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3dgrut:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `3dgrut` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cupy`](/packages/numeric/cupy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`kaolin`](/packages/3d/3dvision/kaolin) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nv-tlabs/3dgrut |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3dgrut)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 3dgrut:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3dgrut)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3dgrut) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3dgrut\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/4k4d/README.md":"# 4k4d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`4k4d:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `4k4d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) [`easyvolcap`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/4k4D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 4k4d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 4k4d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 4k4d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 4k4d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 4k4d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/easyvolcap/README.md":"# easyvolcap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`easyvolcap:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `easyvolcap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/EasyVolcap |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag easyvolcap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host easyvolcap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag easyvolcap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag easyvolcap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build easyvolcap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/fast_gauss/README.md":"# fast_gauss\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fast_gauss:1.0.0`** | |\n| :-- | :-- |\n|    Aliases | `fast_gauss` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/dendenxu/fast-gaussian-rasterization |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fast_gauss)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host fast_gauss:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fast_gauss)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fast_gauss) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fast_gauss\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/gsplat/README.md":"# gsplat\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`gsplat:1.5.3`** | |\n| :-- | :-- |\n|    Aliases | `gsplat` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/gsplat |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag gsplat)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host gsplat:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag gsplat)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag gsplat) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build gsplat\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/nerfview/README.md":"# nerfview\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfview:0.1.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfview` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`tinycudann`](/packages/3d/3dvision/tinycudann) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://pypi.org/project/nerfview/ |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfview)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfview:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfview)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfview) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfview\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/nerf/fruitnerf/README.md":"# fruitnerf\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fruitnerf:1.0`** | |\n| :-- | :-- |\n|    Aliases | `fruitnerf` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio`](/packages/3d/nerf/nerfstudio) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/meyerls/FruitNeRF |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/fruitnerf:r36.3.0`](https://hub.docker.com/r/dustynv/fruitnerf/tags) | `2024-09-02` | `arm64` | `14.6GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fruitnerf)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/fruitnerf:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/fruitnerf:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fruitnerf)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fruitnerf) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fruitnerf\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/nerf/nerfstudio/README.md":"# nerfstudio\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfstudio:1.1.7`** | |\n| :-- | :-- |\n|    Aliases | `nerfstudio` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfstudio |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/nerfstudio:r36.3.0`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2024-09-02` | `arm64` | `7.7GB` |\n|   [`dustynv/nerfstudio:r36.4.0-cu128`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2025-02-11` | `arm64` | `7.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfstudio)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/nerfstudio:r36.4.0-cu128\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/nerfstudio:r36.4.0-cu128\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfstudio)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfstudio) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfstudio\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/cv-cuda/README.md":"# cv-cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cv-cuda:0.15`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15` `cv-cuda` `cvcuda` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg:git`](/packages/multimedia/ffmpeg) [`pyav`](/packages/multimedia/pyav) [`pycuda`](/packages/cuda/pycuda) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cv-cuda:0.15-cpp`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15-cpp` `cv-cuda:cpp` `cvcuda:cpp` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) |\n|    Dependants | [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cv-cuda)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cv-cuda:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cv-cuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cv-cuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cv-cuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/deepstream/README.md":"# deepstream\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`deepstream`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`cmake`](/packages/build/cmake/cmake_pip) [`numpy`](/packages/numeric/numpy) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`gstreamer`](/packages/multimedia/gstreamer) [`tritonserver`](/packages/ml/tritonserver) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-06, 2.3GB)`<br>[`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-22, 6.8GB)`<br>[`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 6.8GB)`<br>[`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-06, 6.7GB)`<br>[`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 9.8GB)` |\n|    Notes | https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_Quickstart.html |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-06` | `arm64` | `2.3GB` |\n|   [`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-22` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-06` | `arm64` | `6.7GB` |\n|   [`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `9.8GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag deepstream)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/deepstream:r32.7.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/deepstream:r32.7.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag deepstream)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag deepstream) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build deepstream\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/3d_diffusion_policy/README.md":"# 3d_diffusion_policy\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n\n3D Diffusion Policy: Generalizable Visuomotor Policy Learning (https://github.com/YanjieZe/3D-Diffusion-Policy)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3d_diffusion_policy`** | |\n| :-- | :-- |\n|    Aliases | `3d-diffusion-policy` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`mujoco`](/packages/sim/mujoco) [`numba`](/packages/numeric/numba) [`diffusers`](/packages/diffusion/diffusers) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) `(2024-09-30, 7.0GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) | `2024-09-30` | `arm64` | `7.0GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3d_diffusion_policy)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/3d_diffusion_policy:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/3d_diffusion_policy:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3d_diffusion_policy)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3d_diffusion_policy) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3d_diffusion_policy\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/ai-toolkit/README.md":"# ai-toolkit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\ndocs.md\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`ai-toolkit`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ai-toolkit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ai-toolkit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ai-toolkit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ai-toolkit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ai-toolkit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/cache_edit/README.md":"# cache_edit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cache_dit:0.2.0`** | |\n| :-- | :-- |\n|    Aliases | `cache_dit` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/vipshop/cache-dit |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cache_edit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cache_edit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cache_edit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cache_edit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cache_edit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/comfyui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fcomfyui%2FREADME.md -->","packages/cv/diffusion/diffusers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusers%2FREADME.md -->","packages/cv/diffusion/diffusion_policy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusion_policy%2FREADME.md -->","packages/cv/diffusion/fastgen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Ffastgen%2FREADME.md -->","packages/cv/diffusion/framepack/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fframepack%2FREADME.md -->","packages/cv/diffusion/sdnext/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fsdnext%2FREADME.md -->","packages/cv/diffusion/self-forcing/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fself-forcing%2FREADME.md -->","packages/cv/diffusion/stable-diffusion-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fstable-diffusion-webui%2FREADME.md -->","packages/cv/holoscan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fholoscan%2FREADME.md -->","packages/cv/jetson-inference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fjetson-inference%2FREADME.md -->","packages/cv/opencv/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fopencv%2FREADME.md -->","packages/cv/vpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fvpi%2FREADME.md -->","packages/hw/canable/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanable%2FREADME.md -->","packages/hw/canbus/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanbus%2FREADME.md -->","packages/hw/cangaroo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcangaroo%2FREADME.md -->","packages/hw/jetcam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjetcam%2FREADME.md -->","packages/hw/jupyter_clickable_image_widget/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjupyter_clickable_image_widget%2FREADME.md -->","packages/hw/librealsense/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Flibrealsense%2FREADME.md -->","packages/hw/oled/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Foled%2FREADME.md -->","packages/hw/pl2303/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fpl2303%2FREADME.md -->","packages/hw/zed/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fzed%2FREADME.md -->","packages/llm/awq/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fawq%2FREADME.md -->","packages/llm/bitsandbytes/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fbitsandbytes%2FREADME.md -->","packages/llm/deepspeed/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2FREADME.md -->","packages/llm/deepspeed/deepspeed-kernels/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2Fdeepspeed-kernels%2FREADME.md -->","packages/llm/dynamo/dynamo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fdynamo%2FREADME.md -->","packages/llm/dynamo/kai-scheduler/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fkai-scheduler%2FREADME.md -->","packages/llm/dynamo/mooncake/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fmooncake%2FREADME.md -->","packages/llm/dynamo/nixl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fnixl%2FREADME.md -->","packages/llm/exllama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fexllama%2FREADME.md -->","packages/llm/gptqmodel/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fgptqmodel%2FREADME.md -->","packages/llm/huggingface_hub/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhuggingface_hub%2FREADME.md -->","packages/llm/hymba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhymba%2FREADME.md -->","packages/llm/ktransformers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fktransformers%2FREADME.md -->","packages/llm/llama-factory/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama-factory%2FREADME.md -->","packages/llm/llama_cpp/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama_cpp%2FREADME.md -->","packages/llm/llamaspeak/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllamaspeak%2FREADME.md -->","packages/llm/lobe_chat/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flobe_chat%2FREADME.md -->","packages/llm/local_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flocal_llm%2FREADME.md -->","packages/llm/minference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fminference%2FREADME.md -->","packages/llm/mistral/mistral_rs/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmistral%2Fmistral_rs%2FREADME.md -->","packages/llm/mlc/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmlc%2FREADME.md -->","packages/llm/nano_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnano_llm%2FREADME.md -->","packages/llm/nemo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnemo%2FREADME.md -->","packages/llm/ollama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Follama%2FREADME.md -->","packages/llm/open-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopen-webui%2FREADME.md -->","packages/llm/openai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopenai%2FREADME.md -->","packages/llm/optimum/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Foptimum%2FREADME.md -->","packages/llm/sglang/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2FREADME.md -->","packages/llm/sglang/genai-bench/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fgenai-bench%2FREADME.md -->","packages/llm/sglang/sgl-kernel/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fsgl-kernel%2FREADME.md -->","packages/llm/sudonim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsudonim%2FREADME.md -->","packages/llm/tensorrt_optimizer/nvidia-modelopt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Fnvidia-modelopt%2FREADME.md -->","packages/llm/tensorrt_optimizer/tensorrt_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Ftensorrt_llm%2FREADME.md -->","packages/llm/text-generation-inference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-inference%2FREADME.md -->","packages/llm/text-generation-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-webui%2FREADME.md -->","packages/llm/transformers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftransformers%2FREADME.md -->","packages/llm/vllm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fvllm%2FREADME.md -->","packages/llm/xgrammar/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fxgrammar%2FREADME.md -->","packages/ml/aim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Faim%2FREADME.md -->","packages/ml/apache/tvm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fapache%2Ftvm%2FREADME.md -->","packages/ml/ctranslate2/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fctranslate2%2FREADME.md -->","packages/ml/dli/dli-nano-ai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fdli%2Fdli-nano-ai%2FREADME.md -->","packages/ml/jax/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fjax%2FREADME.md -->","packages/ml/kans/kat/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fkat%2FREADME.md -->","packages/ml/kans/pykan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fpykan%2FREADME.md -->","packages/ml/l4t/l4t-diffusion/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-diffusion%2FREADME.md -->","packages/ml/l4t/l4t-dynamo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-dynamo%2FREADME.md -->","packages/ml/l4t/l4t-ml/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-ml%2FREADME.md -->","packages/ml/l4t/l4t-pytorch/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-pytorch%2FREADME.md -->","packages/ml/l4t/l4t-tensorflow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-tensorflow%2FREADME.md -->","packages/ml/l4t/l4t-text-generation/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-text-generation%2FREADME.md -->","packages/ml/mamba/causalconv1d/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcausalconv1d%2FREADME.md -->","packages/ml/mamba/cobra/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcobra%2FREADME.md -->","packages/ml/mamba/dimba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fdimba%2FREADME.md -->","packages/ml/mamba/mamba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmamba%2FREADME.md -->","packages/ml/mamba/mambavision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmambavision%2FREADME.md -->","packages/ml/mamba/videomambasuite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fvideomambasuite%2FREADME.md -->","packages/ml/mamba/zigma/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fzigma%2FREADME.md -->","packages/ml/numeric/arrow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Farrow%2FREADME.md -->","packages/ml/numeric/cupy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fcupy%2FREADME.md -->","packages/ml/numeric/numba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumba%2FREADME.md -->","packages/ml/numeric/numpy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumpy%2FREADME.md -->","packages/ml/numeric/warp/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fwarp%2FREADME.md -->","packages/ml/onnx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnx%2FREADME.md -->","packages/ml/onnxruntime/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime%2FREADME.md -->","packages/ml/onnxruntime_genai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime_genai%2FREADME.md -->","packages/ml/pytorch/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2FREADME.md -->","packages/ml/pytorch/apex/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Fapex%2FREADME.md -->","packages/ml/pytorch/torch2trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch2trt%2FREADME.md -->","packages/ml/pytorch/torch3d/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch3d%2FREADME.md -->","packages/ml/pytorch/torch_tensorrt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch_tensorrt%2FREADME.md -->","packages/ml/pytorch/torchao/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchao%2FREADME.md -->","packages/ml/pytorch/torchaudio/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchaudio%2FREADME.md -->","packages/ml/pytorch/torchsaver/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsaver%2FREADME.md -->","packages/ml/pytorch/torchsde/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsde%2FREADME.md -->","packages/ml/pytorch/torchtext/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchtext%2FREADME.md -->","packages/ml/pytorch/torchvision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchvision%2FREADME.md -->","packages/ml/rapids/cudf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcudf%2FREADME.md -->","packages/ml/rapids/cuml/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcuml%2FREADME.md -->","packages/ml/rapids/raft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fraft%2FREADME.md -->","packages/ml/tensorflow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2FREADME.md -->","packages/ml/tensorflow/graphics/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Fgraphics%2FREADME.md -->","packages/ml/tensorflow/tensorboard/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftensorboard%2FREADME.md -->","packages/ml/tensorflow/text/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftext%2FREADME.md -->","packages/ml/transformer-engine/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftransformer-engine%2FREADME.md -->","packages/ml/triton/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftriton%2FREADME.md -->","packages/ml/tritonserver/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftritonserver%2FREADME.md -->","packages/ml/xlstm/mlstm_kernels/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fmlstm_kernels%2FREADME.md -->","packages/ml/xlstm/pltsm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fpltsm%2FREADME.md -->","packages/ml/xlstm/xlstm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fxlstm%2FREADME.md -->","packages/multimedia/decord/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fdecord%2FREADME.md -->","packages/multimedia/ffmpeg/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fffmpeg%2FREADME.md -->","packages/multimedia/gstreamer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fgstreamer%2FREADME.md -->","packages/multimedia/jetson-utils/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fjetson-utils%2FREADME.md -->","packages/multimedia/libcom/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Flibcom%2FREADME.md -->","packages/multimedia/opengl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fopengl%2FREADME.md -->","packages/multimedia/pyav/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fpyav%2FREADME.md -->","packages/multimedia/sound-utils/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fsound-utils%2FREADME.md -->","packages/multimedia/video-codec-sdk/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvideo-codec-sdk%2FREADME.md -->","packages/multimedia/vulkan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvulkan%2FREADME.md -->","packages/net/devpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fdevpi%2FREADME.md -->","packages/net/pypi.dev/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fpypi.dev%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-policy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-policy%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-predict2/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-predict2%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-reason1/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-reason1%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-transfer1/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-transfer1%2FREADME.md -->","packages/physicalAI/cosmos/cosmos_diffusion_renderer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2FREADME.md -->","packages/physicalAI/cosmos/cosmos_diffusion_renderer/nvdiffrast/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2Fnvdiffrast%2FREADME.md -->","packages/physicalAI/isaac-ros/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fisaac-ros%2FREADME.md -->","packages/physicalAI/lerobot/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Flerobot%2FREADME.md -->","packages/physicalAI/opendronemap/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopendronemap%2FREADME.md -->","packages/physicalAI/openpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopenpi%2FREADME.md -->","packages/physicalAI/protomotions/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fprotomotions%2FREADME.md -->","packages/physicalAI/ros/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fros%2FREADME.md -->","packages/physicalAI/sim/genesis/Readme.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FReadme.md -->","packages/physicalAI/sim/genesis/quadrants/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fquadrants%2FREADME.md -->","packages/physicalAI/sim/genesis/splashSurf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FsplashSurf%2FREADME.md -->","packages/physicalAI/sim/genesis/vtk/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fvtk%2FREADME.md -->","packages/physicalAI/sim/habitat-sim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fhabitat-sim%2FREADME.md -->","packages/physicalAI/sim/isaac-sim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2FREADME.md -->","packages/physicalAI/sim/isaac-sim/isaac-lab/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2Fisaac-lab%2FREADME.md -->","packages/physicalAI/sim/mimicgen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmimicgen%2FREADME.md -->","packages/physicalAI/sim/mujoco/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmujoco%2FREADME.md -->","packages/physicalAI/sim/newton/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fnewton%2FREADME.md -->","packages/physicalAI/sim/physx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fphysx%2FREADME.md -->","packages/physicalAI/sim/robogen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobogen%2FREADME.md -->","packages/physicalAI/sim/robomimic/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobomimic%2FREADME.md -->","packages/physicalAI/sim/robosuite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobosuite%2FREADME.md -->","packages/physicalAI/sim/sapien/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fsapien%2FREADME.md -->","packages/physicalAI/vla/crossformer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fcrossformer%2FREADME.md -->","packages/physicalAI/vla/isaac-gr00t/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fisaac-gr00t%2FREADME.md -->","packages/physicalAI/vla/octo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Focto%2FREADME.md -->","packages/physicalAI/vla/openvla/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fopenvla%2FREADME.md -->","packages/physicalAI/vla/robopoint/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Frobopoint%2FREADME.md -->","packages/rag/graphiti/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fgraphiti%2FREADME.md -->","packages/rag/jetson-copilot/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fjetson-copilot%2FREADME.md -->","packages/rag/langchain/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Flangchain%2FREADME.md -->","packages/rag/llama-index/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fllama-index%2FREADME.md -->","packages/rag/memvid/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fmemvid%2FREADME.md -->","packages/smart-home/dependencies/ciso8601/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fciso8601%2FREADME.md -->","packages/smart-home/dependencies/psutil-home-assistant/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fpsutil-home-assistant%2FREADME.md -->","packages/smart-home/homeassistant-base/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-base%2FREADME.md -->","packages/smart-home/homeassistant-core/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-core%2FREADME.md -->","packages/smart-home/wyoming/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2FREADME.md -->","packages/smart-home/wyoming/wyoming-assist-microphone/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-assist-microphone%2FREADME.md -->","packages/smart-home/wyoming/wyoming-openwakeword/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-openwakeword%2FREADME.md -->","packages/smart-home/wyoming/wyoming-piper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-piper%2FREADME.md -->","packages/smart-home/wyoming/wyoming-whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-whisper%2FREADME.md -->","packages/speech/audiocraft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Faudiocraft%2FREADME.md -->","packages/speech/chatterbox-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fchatterbox-tts%2FREADME.md -->","packages/speech/espeak/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fespeak%2FREADME.md -->","packages/speech/faster-whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Ffaster-whisper%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-fastapi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-fastapi%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-hf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-hf%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-onnx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-onnx%2FREADME.md -->","packages/speech/piper1-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fpiper1-tts%2FREADME.md -->","packages/speech/riva-client/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Friva-client%2FREADME.md -->","packages/speech/spark-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspark-tts%2FREADME.md -->","packages/speech/speaches/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeaches%2FREADME.md -->","packages/speech/speech-dispatcher/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeech-dispatcher%2FREADME.md -->","packages/speech/voice-pro/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoice-pro%2FREADME.md -->","packages/speech/voicecraft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoicecraft%2FREADME.md -->","packages/speech/whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper%2FREADME.md -->","packages/speech/whisper_trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper_trt%2FREADME.md -->","packages/speech/whisperx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisperx%2FREADME.md -->","packages/speech/xtts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fxtts%2FREADME.md -->","packages/vectordb/faiss/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss%2FREADME.md -->","packages/vectordb/faiss_lite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss_lite%2FREADME.md -->","packages/vectordb/nanodb/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Fnanodb%2FREADME.md -->","packages/vit/clip_trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fclip_trt%2FREADME.md -->","packages/vit/efficientvit/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fefficientvit%2FREADME.md -->","packages/vit/nanoowl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanoowl%2FREADME.md -->","packages/vit/nanosam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanosam%2FREADME.md -->","packages/vit/sam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsam%2FREADME.md -->","packages/vit/sapiens/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsapiens%2FREADME.md -->","packages/vit/tam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Ftam%2FREADME.md -->","packages/vlm/lita/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Flita%2FREADME.md -->","packages/vlm/llama-vision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllama-vision%2FREADME.md -->","packages/vlm/llava/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllava%2FREADME.md -->","packages/vlm/minigpt4/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fminigpt4%2FREADME.md -->","packages/vlm/prismatic/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fprismatic%2FREADME.md -->","packages/vlm/videollama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvideollama%2FREADME.md -->","packages/vlm/vila-microservice/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila-microservice%2FREADME.md -->","packages/vlm/vila/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila%2FREADME.md -->","packages/vlm/xtuner/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fxtuner%2FREADME.md -->"},"files":{".claude/skills/jetson-pr.md":"# Create a jetson-containers Pull Request\n\nHelp the user create a well-formed PR that follows this repository's conventions.\n\n## Repository PR conventions\n\n- **Target branch**: PRs go to `dev`, not `master`. The CI workflow `pr-on-dev.yml` runs self-hosted Jetson builds against `dev`.\n- **Scope**: each PR should touch only the packages it claims to change. Keep unrelated cleanup out.\n- **Commit style**: short, lowercase, imperative subject. No period. Examples from git log: `fix`, `cuda 13.2 needs newer drivers to run in Orin`, `tensorrt compatible orin`, `Update ffmpeg version from 8.1 to 8.0.1`.\n- **No force-push** to shared branches.\n\n## Step-by-step workflow\n\n1. **Understand the changes**: run `git diff master...HEAD` and `git log master..HEAD --oneline` to see exactly what is changing. Read the changed files if needed to understand the intent.\n\n2. **Identify affected packages**: map every changed file back to its package directory (`packages/<category>/<name>/`). List them — this drives the PR title and body.\n\n3. **Stage and commit** (if there are uncommitted changes):\n - Stage only the files related to this PR.\n - Write a commit message following the style above. No trailing period.\n - Do not skip hooks (`--no-verify`).\n\n4. **Create or switch to a feature branch** if on `master`:\n ```\n git checkout -b <short-descriptive-name>\n ```\n Branch name: lowercase, hyphenated, describes the change (e.g. `upgrade-pytorch-2.9`, `add-sglang-0.4`, `fix-tensorrt-orin`).\n\n5. **Push and open the PR**:\n ```\n gh pr create --base dev --title \"<title>\" --body \"<body>\"\n ```\n\n6. **PR title format**: `<verb> <package(s)> <version or brief description>`\n - Version bumps: `upgrade pytorch to 2.9`\n - New packages: `add sglang 0.4`\n - Bug fixes: `fix tensorrt build on Orin`\n - Multi-package: list up to 3, then `and N others`\n\n7. **PR body template**:\n ```\n ## Changes\n - <package>: <what changed and why>\n\n ## Packages affected\n - `packages/<category>/<name>/`\n\n ## Test plan\n - [ ] `jetson-containers build <package>` on Orin (JP6)\n - [ ] Add any version-specific test steps here\n\n ## Notes\n <Any JetPack version constraints, driver requirements, or upstream release links>\n ```\n\n8. **Return the PR URL** when done.\n\n## What good PRs look like in this repo\n\n- Version upgrades include the new version entry in `config.py` **and** update `version.py` if it exists.\n- New packages include a `Dockerfile`, metadata (YAML header or `config.yaml`), and at minimum a stub `test.py` or `test.sh`.\n- Dependency bumps update `requires` constraints if the new version drops support for an older JetPack.\n- PRs do not mix version bumps with unrelated refactors.\n\n## Before opening the PR, verify\n\n- `git diff --stat dev...HEAD` shows only the intended files.\n- No `.env`, secrets, or build artifacts are staged.\n- The branch name and title are consistent.\n",".claude/skills/upgrade-package.md":"# Upgrade Package Version\n\nHelp the user upgrade the version of one or more packages in this jetson-containers repository.\n\n## What the user will provide\n\nThe user will tell you the package name and the new version. If they only give you the package name, ask for the target version before proceeding.\n\n## How version definitions work\n\nEach package under `packages/<category>/<name>/` defines its version in one of these ways:\n\n1. **`version.py`** — a standalone file that exports a version constant, e.g. `PYTORCH_VERSION = Version('2.8')`. This is the canonical version used by `config.py`.\n2. **`config.py`** — calls a helper function (like `pytorch_pip(version, requires=...)`) with explicit version strings in a `package = [...]` list. Adding a new version = adding a new line to this list.\n3. **`config.yaml` / YAML header in Dockerfile** — static `build_args` with version pinned directly.\n4. **`Dockerfile` `ARG` lines** — `ARG SOME_VERSION=x.y.z`.\n\n## Step-by-step workflow\n\n1. **Locate the package directory**: search under `packages/` for the package name.\n2. **Read all config files** in that directory (`config.py`, `config.yaml`, `version.py`, `Dockerfile`, `Dockerfile.*`).\n3. **Identify where the version is defined** (see above). Show the user the current version(s) before making changes.\n4. **Apply the upgrade**:\n - If there is a `version.py`, update the version constant there first; `config.py` will pick it up automatically.\n - If `config.py` uses a list of `pytorch_pip(version, ...)` calls, **add** a new entry for the new version (do not remove old ones — the repo keeps historical versions for different JetPack targets). Set the new version as the latest by updating the `PYTORCH_VERSION`-like constant in `version.py`.\n - If the version is in a YAML `build_args` or a Dockerfile `ARG`, do a targeted edit of that value.\n5. **Check for cascading dependencies**: grep for the old version string across other config files in the same package directory. Common locations: `Dockerfile` `ARG` defaults, `build_args` dicts, `requires` strings, download URLs, wheel filenames.\n6. **Update download artifacts if needed**: if the Dockerfile fetches a wheel, tarball, or pre-built binary, update the URL/filename to match the new version. Check the upstream source (PyPI, NVIDIA, GitHub releases) if the user hasn't provided a URL.\n7. **Show a diff summary** of every file you changed before finishing.\n\n## Things to preserve\n\n- Do **not** delete existing version entries in `config.py` lists — old JetPack targets still need them.\n- Keep the `requires` constraints on old entries unchanged.\n- Do not change `alias` lists for older versions.\n- Follow the existing code style (no reformatting unrelated code, no added comments).\n\n## When to ask vs. proceed\n\n- Ask for the target version if not given.\n- Ask about JetPack `requires` constraint only if the new version introduces a minimum JetPack requirement that differs from the previous latest entry.\n- Proceed with everything else silently and show the diff at the end.\n","AGENTS.md":"# CLAUDE.md\n\nThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.\n\n## Overview\n\njetson-containers is a modular container build system for NVIDIA Jetson devices. It discovers packages, resolves their dependencies, chains their Dockerfiles together, and runs `docker build` to produce multi-stage images optimized for specific L4T/JetPack/CUDA versions.\n\n## Commands\n\n### Build & Run\n```bash\n# Build one or more packages (chains Dockerfiles in dependency order)\njetson-containers build <package> [<package2> ...]\n# or equivalently:\n./build.sh <package> [<package2> ...]\n\n# Find and run a compatible container image\njetson-containers run $(autotag <package>) [command]\n# or equivalently:\n./run.sh $(autotag <package>) [command]\n\n# Find the best matching image tag (local → DockerHub → build)\nautotag <package>\n\n# Useful build flags\njetson-containers build --simulate <package> # dry-run: print commands without building\njetson-containers build --multiple pytorch tensorflow # build each package as a separate image\njetson-containers build --list-packages # list all discovered packages\njetson-containers build --show-packages <package> # show resolved metadata for a package\njetson-containers build --base=xyz:latest <package> # start the build chain from an existing image\njetson-containers build --push=myuser <package> # push to a registry after building\n\n# Check for upstream version upgrades (--apply to write changes)\njetson-containers upgrade [<package> ...] [--apply]\n```\n\n### Code Style\n```bash\n# Install pre-commit hooks (one-time setup)\npip install -r requirements.txt\npre-commit install\n\n# Run linting/formatting checks manually\npre-commit run --all-files\n```\n\nPre-commit runs **Black** (formatter, line-length=88) and **Flake8** (linter) only on `jetson_containers/` and `test_precommit.py`. `packages/` Dockerfiles and `config.py` files are **not** linted by pre-commit. Note: `skip-string-normalization = true` in `pyproject.toml`.\n\n### Tests\n```bash\n# Run a package config unit test directly (uses stdlib unittest, no pytest needed)\npython3 tests/verify_triton_config.py\n```\n\n### Installation\n```bash\nbash install.sh # sets up venv, installs deps, symlinks CLI tools to /usr/local/bin\n```\n\n## Architecture\n\n### Package System\n\nEvery buildable unit is a **package**. Packages live under `packages/<category>/<name>/` and define metadata via one of (in priority order):\n\n1. **YAML header in Dockerfile** — metadata between `#---` markers at the top\n2. **`config.yaml` / `config.yml`** — static YAML config file\n3. **`config.json`** — meta-container (no Dockerfile, just dependency composition)\n4. **`config.py`** — dynamic config, returns a list of package dicts at build time\n\nKey metadata fields: `name`, `alias`, `depends` (build-order deps), `requires` (version constraints on L4T/CUDA/Python), `test`, `build_args`, `dockerfile`.\n\n### Build Pipeline\n\n1. **`jetson_containers/packages.py`** — scans `packages/` recursively, runs `config.py` scripts, resolves the dependency graph, filters by L4T version compatibility, and exposes a custom Python meta path finder so packages can `from packages.ml.pytorch.version import ...`\n2. **`jetson_containers/build.py`** — CLI entry point; parses args and dispatches\n3. **`jetson_containers/container.py`** — core build logic: stitches multiple package Dockerfiles into one combined `Dockerfile`, invokes `docker build`, optionally runs per-package tests, and optionally pushes to a registry\n4. **`jetson_containers/l4t_version.py`** — detects architecture (tegra-aarch64, aarch64, x86_64) and reads L4T/JetPack/CUDA/GPU-arch info from `/etc/nv_tegra_release`; all version variables (e.g. `CUDA_VERSION`, `PYTORCH_VERSION`) flow from here into build args\n\n### Dynamic Package Generation\n\n`config.py` files execute at build time and can return multiple package dicts to produce version variants. For example, PyTorch's `config.py` generates `pytorch:2.8`, `pytorch:2.8-all`, `pytorch:2.8-builder`, etc. Aliases (`torch` → `pytorch:2.8`) are resolved during package discovery.\n\nWhen a package's version selection logic is complex, put it in a separate `version.py` alongside `config.py`. This lets other packages cross-import it:\n```python\nfrom packages.ml.pytorch.version import PYTORCH_VERSION\n```\nThe `_PackagesFinder` meta path finder in `packages.py` resolves these imports against the `packages/` tree without requiring `__init__.py` files.\n\n### Dockerfile Conventions\n\nUse `uv pip install` (not plain `pip install`) inside Dockerfiles — `uv` is pre-installed in the base images and is significantly faster. Every Dockerfile must start with:\n```dockerfile\nARG BASE_IMAGE\nFROM ${BASE_IMAGE}\n```\n\n### Version Overrides\n\nCopy `.env.default` → `.env` to override `L4T_VERSION`, `CUDA_VERSION`, `LSB_RELEASE`, `CUDA_ARCH`, local PyPI/APT mirror URLs, and SCP/webhook credentials without modifying tracked files.\n\n### Creating a New Package\n\n1. Create `packages/<category>/<name>/Dockerfile` with a `#---` YAML header (or a `config.yaml`/`config.py`).\n2. Declare `depends` on upstream packages (e.g., `pytorch`, `cuda`).\n3. Optionally add `test.py` or `test.sh` for post-build validation.\n4. Build: `jetson-containers build <name>`\n\n## Security Features\n\n### Multi-Stage Builds\n\nThe build system is multi-stage aware. Packages can use `FROM ... AS <alias>` for intermediate stages (e.g. a compiler/builder stage) without those stages polluting the final image. The build system tracks stage boundaries and only injects BuildKit features (ccache env vars, `--device` mounts, ccache `--mount` flags) into the **final** stage (the `FROM ${BASE_IMAGE}` stage). Intermediate stages are left untouched.\n\nDeclare multi-stage packages in config metadata:\n```yaml\n#---\n# name: mypackage\n# multistage: true\n# depends: [cuda]\n#---\nARG BASE_IMAGE\nFROM gcc:12 AS builder\nRUN make install\n\nFROM ${BASE_IMAGE}\nCOPY --from=builder /usr/local /usr/local\nRUN ldconfig\n```\n\nThe final stage **must** use `ARG BASE_IMAGE` / `FROM ${BASE_IMAGE}` so it chains correctly in the dependency build pipeline.\n\n### Non-Root User\n\n**At build time** — pass identity via env vars; packages can create and switch to the declared user:\n```bash\nexport CONTAINER_USER=jetson\nexport CONTAINER_UID=1000\nexport CONTAINER_GID=1000\njetson-containers build mypackage\n```\n\n`CONTAINER_USER`, `CONTAINER_UID`, `CONTAINER_GID` are forwarded as `--build-arg` automatically. Packages consume them like:\n```dockerfile\nARG CONTAINER_USER=root\nARG CONTAINER_UID=0\nARG CONTAINER_GID=0\nRUN if [ \"$CONTAINER_UID\" != \"0\" ]; then \\\n groupadd -g $CONTAINER_GID $CONTAINER_USER && \\\n useradd -u $CONTAINER_UID -g $CONTAINER_GID -m $CONTAINER_USER; \\\n fi\nUSER $CONTAINER_USER\n```\n\nDeclare the supported user in package metadata (`run_user` is informational for tooling):\n```yaml\n# run_user: jetson\n```\n\n**At runtime** — set `DOCKER_USER` to run the container as a non-root identity:\n```bash\nDOCKER_USER=1000:1000 ./run.sh myimage\n# or by name (the user must exist in the image):\nDOCKER_USER=jetson ./run.sh myimage\n```\nX11 forwarding (`xhost`) is automatically updated to grant the declared user display access.\n\nFor GPU access, `run.sh` also discovers the numeric GIDs that own the NVIDIA/Tegra device nodes (`/dev/nvhost*`, `/dev/nvmap`, `/dev/nvgpu/*`, `/dev/dri/render*`) and adds them via `--group-add <gid>`. Without this, a non-root process can't open the GPU nodes and CUDA fails with error 801 (`cudaGetDeviceCount` \"operation not supported\"). Numeric GIDs are used because the host's `render` GID usually doesn't match the container's. See [`docs/run.md`](/docs/run.md#gpu-access-as-non-root-cuda-error-801).\n\n### Secrets\n\n**Build-time secrets** (never baked into image layers) — declare secrets in package metadata:\n```yaml\n#---\n# name: mypackage\n# secrets: [huggingface_token]\n#---\n```\n\nPoint each secret at a file on the host via `JETSON_SECRET_<NAME>`:\n```bash\nexport JETSON_SECRET_HUGGINGFACE_TOKEN=~/.hf_token\njetson-containers build mypackage\n```\n\nSecrets are mounted as `/run/secrets/<name>` inside `RUN` steps only (BuildKit `--secret`). Use them in the Dockerfile:\n```dockerfile\nRUN --mount=type=secret,id=huggingface_token \\\n HF_TOKEN=$(cat /run/secrets/huggingface_token) \\\n python3 download_model.py\n```\n\nIf the env var is missing, a warning is printed and the build continues without that secret.\n\n**Runtime secrets** — prefer file-backed tokens over env vars (values in env vars are visible in `docker inspect`):\n```bash\n# Preferred: mount token file into /run/secrets/\nHUGGINGFACE_TOKEN_FILE=~/.hf_token ./run.sh myimage\n\n# Fallback (token visible in docker inspect):\nHUGGINGFACE_TOKEN=hf_xxx ./run.sh myimage\n```\n\n`HUGGINGFACE_TOKEN_FILE` mounts the file read-only at `/run/secrets/huggingface_token` and sets `HF_TOKEN_FILE` inside the container.\n\n### Key Supporting Modules\n\n- `jetson_containers/tag.py` — image tagging conventions\n- `jetson_containers/docs.py` — auto-generates per-package docs from metadata\n- `jetson_containers/ci.py` — CI/CD helpers\n- `jetson_containers/network.py` / `webhook.py` — GitHub API integration and build notifications\n- `jetson_containers/logging.py` — colored terminal output used throughout\n- `jetson_containers/upgrade.py` — checks packages against upstream (PyPI/GitHub) for newer versions; powers `jetson-containers upgrade`\n- `jetson_containers/db.py` — syncs DockerHub/GitHub metadata for registry introspection\n","CLAUDE.md":"@AGENTS.md\n","packages/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FREADME.md -->","packages/attention/ParaAttention/README.md":"# ParaAttention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nhttps://github.com/chengzeyi/ParaAttention\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`paraattention:0.4.0`** | |\n| :-- | :-- |\n|    Aliases | `paraattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ParaAttention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ParaAttention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ParaAttention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ParaAttention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ParaAttention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/block-sparse-attention/README.md":"# block-sparse-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`block_sparse_attn:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `block_sparse_attn` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`xattention:0.0.1`](/packages/attention/xattention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag block-sparse-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host block-sparse-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag block-sparse-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag block-sparse-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build block-sparse-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flash-attention/README.md":"# flash-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flash-attention:2.5.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.6.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.4.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.0.post2`** | |\n| :-- | :-- |\n|    Aliases | `flash-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flashinfer/README.md":"# flash-infer\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flashinfer:0.2.1.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:0.8.4`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.6.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`mlc:0.20.0`](/packages/llm/mlc) [`sglang:0.4.6`](/packages/llm/sglang) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.8`** | |\n| :-- | :-- |\n|    Aliases | `flashinfer` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`llama-factory`](/packages/llm/llama-factory) [`local_llm`](/packages/llm/local_llm) [`mlc:0.21.0`](/packages/llm/mlc) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-infer)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-infer:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-infer)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-infer) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-infer\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/flexprefill/README.md":"# flexprefill\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flexprefill:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `flexprefill` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`llama-factory`](/packages/llm/llama-factory) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`sdnext`](/packages/diffusion/sdnext) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flexprefill)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flexprefill:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flexprefill)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flexprefill) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flexprefill\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/huggingface_kernels/README.md":"# huggingface_kernels\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`huggingface_kernels:0.7.0`** | |\n| :-- | :-- |\n|    Aliases | `huggingface_kernels` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`huggingface_hub`](/packages/llm/huggingface_hub) [`torchvision`](/packages/pytorch/torchvision) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag huggingface_kernels)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host huggingface_kernels:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag huggingface_kernels)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag huggingface_kernels) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build huggingface_kernels\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/log-linear-attention/README.md":"# log-linear-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`log-linear-attention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `log-linear-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`ninja`](/packages/build/ninja) [`torchaudio`](/packages/pytorch/torchaudio) [`causalconv1d`](/packages/ml/mamba/causalconv1d) [`mamba`](/packages/ml/mamba/mamba) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag log-linear-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host log-linear-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag log-linear-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag log-linear-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build log-linear-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/radial-attention/README.md":"# radial-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`radial-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `radial-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`flashinfer`](/packages/attention/flash-infer) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag radial-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host radial-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag radial-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag radial-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build radial-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/sage-attention/README.md":"# sage-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sage-attention:3.0.0`** | |\n| :-- | :-- |\n|    Aliases | `sage-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`framepack`](/packages/diffusion/framepack) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sage-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sage-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sage-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sage-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sage-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/sparge-attention/README.md":"# sparge-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparge-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparge-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/thu-ml/SpargeAttn.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparge-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparge-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparge-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparge-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparge-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/xattention/README.md":"# xattention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xattention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `xattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`block_sparse_attn`](/packages/attention/block-sparse-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xattention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host xattention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xattention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xattention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xattention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/attention/xformers/README.md":"# xformers\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xformers:0.0.32`** | |\n| :-- | :-- |\n|    Aliases | `xformers` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dependants | [`apex:0.1`](/packages/pytorch/apex) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n| **`xformers:0.0.33`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/xformers:r35.2.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-06` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.3.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-14` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.4.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2024-01-09` | `arm64` | `5.9GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xformers)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/xformers:r35.4.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/xformers:r35.4.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xformers)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xformers) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xformers\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/code/jupyterlab/README.md":"# jupyterlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nRunning the `jupyerlab` container will automatically start a JupyterLab server in the background on port 8888, with default login password `nvidia`. The JupyterLab server logs will be saved to `/data/logs/jupyter.log` if you need to inspect them (this location is automatically mounted under your `jetson-containers/data` directory)\n\nTo change the default settings, you can set the `$JUPYTER_ROOT`, `$JUPYTER_PORT`, `$JUPYTER_PASSWORD`, and `$JUPYTER_LOG` environment variables when starting the container like so:\n\n```bash\njetson-containers run \\\n --env JUPYTER_ROOT=/home/user \\\n --env JUPYTER_PORT=8000 \\\n --env JUPYTER_PASSWORD=password \\\n --env JUPYTER_LOGS=/dev/null \\\n $(autotag jupyterlab)\n```\n\nThe [`/start_jupyter`](./start_jupyter) script is the default CMD that the container runs when it starts - however, if you don't want the JupyterLab server started by default, you can either add a different CMD in your own Dockerfile, or override it at startup:\n\n```bash\n# skip straight to the terminal instead of starting JupyterLab first\njetson-containers run /bin/bash\n```\n\nYou can then still manually run the [`/start_jupyter`](./start_jupyter) script later when desired.\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`jupyterlab:latest`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:latest-myst`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab:myst` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:latest`](/packages/code/jupyterlab) |\n|    Dependants | [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0-myst`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/jupyterlab:r32.7.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.7GB` |\n|   [`dustynv/jupyterlab:r35.2.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-12-06` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r35.3.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `5.4GB` |\n|   [`dustynv/jupyterlab:r35.4.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-10-07` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r36.2.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.6GB` |\n|   [`dustynv/jupyterlab:r36.4.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-10` | `arm64` | `5.8GB` |\n|   [`dustynv/jupyterlab:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-03` | `arm64` | `5.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag jupyterlab)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/jupyterlab:r36.4.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/jupyterlab:r36.4.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag jupyterlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag jupyterlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build jupyterlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/code/vscode/README.md":"# vscode\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`vscode`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:cuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`tensorrt`](/packages/cuda/tensorrt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:torch`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`tensorrt`](/packages/cuda/tensorrt) [`torch2trt`](/packages/pytorch/torch2trt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:transformers`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`bitsandbytes`](/packages/llm/bitsandbytes) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag vscode)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host vscode:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag vscode)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag vscode) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build vscode\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/cuda-python/README.md":"# cuda-python\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda-python:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda-python` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cuml`](/packages/ml/rapids/cuml) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss_lite`](/packages/vectordb/faiss_lite) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index:samples`](/packages/rag/llama-index) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`raft`](/packages/ml/rapids/raft) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`tam`](/packages/vit/tam) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.8`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu128`](/packages/cuda/cuda) [`cuda:12.8`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.9`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu129`](/packages/cuda/cuda) [`cuda:12.9`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:13.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu130`](/packages/cuda/cuda) [`cuda:13.0`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/cuda-python:builder-r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-26` | `arm64` | `5.1GB` |\n|   [`dustynv/cuda-python:builder-r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-11` | `arm64` | `3.6GB` |\n|   [`dustynv/cuda-python:r35.2.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.3.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/cuda-python:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cuda-python)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/cuda-python:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/cuda-python:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cuda-python)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cuda-python) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cuda-python\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/cuda/README.md":"# cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/cuda:12.2-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-05, 3.4GB)`<br>[`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n\n| **`pip_cache:cu122`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2`](/packages/cuda/cuda) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.2-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Images | [`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu124`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4`](/packages/cuda/cuda) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.4-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda` |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`clip_trt`](/packages/vit/clip_trt) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pyav`](/packages/multimedia/pyav) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/packages/speech/whisperx) [`wyoming-piper:1.6.2`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-piper:master`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-whisper:2.5.0`](/packages/smart-home/wyoming/wyoming-whisper) [`wyoming-whisper:master`](/packages/smart-home/wyoming/wyoming-whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xformers:0.0.32`](/packages/attention/xformers) [`xformers:0.0.33`](/packages/attention/xformers) [`xgrammar:0.1.15`](/packages/llm/xgrammar) [`xgrammar:0.1.18`](/packages/llm/xgrammar) [`xgrammar:0.1.19`](/packages/llm/xgrammar) [`xgrammar:0.1.20`](/packages/llm/xgrammar) [`xgrammar:0.1.21`](/packages/llm/xgrammar) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtts`](/packages/speech/xtts) [`xtuner`](/packages/vlm/xtuner) [`zed:5.0`](/packages/hw/zed) [`zed:5.0-humble`](/packages/hw/zed) [`zed:5.0-jazzy`](/packages/hw/zed) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu126`** | |\n| :-- | :-- |\n|    Aliases | `pip_cache` `pip-cache` |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`aim`](/packages/ml/aim) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`canbus`](/packages/hw/canbus) [`cangaroo`](/packages/hw/cangaroo) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`ciso8601:v2.3.2`](/packages/smart-home/dependencies/ciso8601) [`clip_trt`](/packages/vit/clip_trt) [`cmake:pip`](/packages/build/cmake/cmake_pip) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6`](/packages/cuda/cuda) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`docker`](/packages/build/docker) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`graphiti`](/packages/rag/graphiti) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`h5py`](/packages/build/h5py) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`homeassistant-core:2025.7.0`](/packages/smart-home/homeassistant-core) [`huggingface_hub`](/packages/llm/huggingface_hub) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:python`](/packages/multimedia/jetson-utils) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llamaspeak`](/packages/llm/llamaspeak) [`llava`](/packages/vlm/llava) [`llvm:10`](/packages/build/llvm) [`llvm:11`](/packages/build/llvm) [`llvm:12`](/packages/build/llvm) [`llvm:13`](/packages/build/llvm) [`llvm:14`](/packages/build/llvm) [`llvm:15`](/packages/build/llvm) [`llvm:16`](/packages/build/llvm) [`llvm:17`](/packages/build/llvm) [`llvm:18`](/packages/build/llvm) [`llvm:19`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`meshlab:MeshLab-2023.12`](/packages/3d/3dvision/meshlab) [`meshlab:MeshLab-2025.03`](/packages/3d/3dvision/meshlab) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`mujoco`](/packages/sim/mujoco) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`ninja`](/packages/build/ninja) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`numpy`](/packages/numeric/numpy) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnx`](/packages/ml/onnx) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`openai`](/packages/llm/openai) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`polyscope:2.5.0`](/packages/3d/3dvision/polyscope) [`prismatic`](/packages/vlm/prismatic) [`protobuf:apt`](/packages/build/protobuf/protobuf_apt) [`protobuf:cpp`](/packages/build/protobuf/protobuf_cpp) [`protomotions:2.5.0`](/packages/robots/protomotions) [`psutil-home-assistant:0.0.1`](/packages/smart-home/dependencies/psutil-home-assistant) [`pyav`](/packages/multimedia/pyav) [`pybind11`](/packages/build/pybind11) [`pyceres:2.5`](/packages/3d/3dvision/pyceres) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`python:3.10`](/packages/build/python) [`python:3.11`](/packages/build/python) [`python:3.12`](/packages/build/python) [`python:3.13`](/packages/build/python) [`python:3.14`](/packages/build/python) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`riva-client:python`](/packages/speech/riva-client) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`rust`](/packages/build/rust) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim`](/packages/llm/sudonim) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorboard`](/packages/ml/tensorflow/tensorboard) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode`](/packages/code/vscode) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`vtk:9.3.1`](/packages/sim/genesis/vtk) [`vtk:9.4.2`](/packages/sim/genesis/vtk) [`vtk:9.5.0`](/packages/sim/genesis/vtk) [`vulkan:1.3.275`](/packages/multimedia/vulkan) [`vulkan:1.3.296`](/packages/multimedia/vulkan) [`vulkan:1.4.309`](/packages/multimedia/vulkan) [`vulkan:1.4.313`](/packages/multimedia/vulkan) [`vulkan:1.4.314`](/packages/multimedia/vulkan) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/pa\n\n<!-- [TRUNCATED AT 50KB] -->","packages/cuda/cudastack/README.md":"# CUDA Stack - Consolidated Library Installation\n\n## What is cudastack?\n\n`cudastack` is a **consolidation package** that installs multiple CUDA libraries (cuDNN, NCCL, TensorRT, CUTLASS, etc.) in **ONE RUN command** to avoid Docker's \"max depth exceeded\" error.\n\n## Important: cudastack vs cuda\n\n```\ncuda/ cudastack/\n├── Dockerfile ├── Dockerfile (ONE RUN for all libs)\n├── Dockerfile.pip ├── config.py (variants)\n├── Dockerfile.builtin └── install/ & build/ scripts\n├── Dockerfile.samples\n└── config.py\n ├── cuda_package() cudastack DEPENDS ON cuda\n ├── cuda_builtin() (doesn't replace it!)\n ├── cuda_samples()\n └── pip_cache()\n```\n\n**Key Points:**\n\n1. ✅ **`cuda` package** - Installs CUDA toolkit, provides pip_cache, samples, builtin variants\n2. ✅ **`cudastack` package** - Builds ON TOP of `cuda`, adds cuDNN/NCCL/TensorRT/etc. in ONE layer\n3. ✅ **pip_cache, cuda_samples, cuda_builtin** - Still work! They're part of `cuda`, not `cudastack`\n\n## The Problem It Solves\n\n### Before (Separate Packages):\n```\nLayer 1: cuda\nLayer 2: cudnn\nLayer 3: tensorrt\nLayer 4: nccl\nLayer 5: cutlass\n...\nLayer 50+: \"max depth exceeded\" ❌\n```\n\n### After (cudastack):\n```\nLayer 1: cuda (with pip_cache, samples, etc.)\nLayer 2: cudastack (cudnn + nccl + tensorrt + cutlass in ONE RUN)\nLayer 3: your-app\nResult: Works! ✅\n```\n\n## Architecture\n\n```\nBASE_IMAGE\n ↓\ncuda:12.4 (from cuda/Dockerfile)\n ├── pip_cache:cu124 (from cuda/Dockerfile.pip)\n └── cuda:12.4-samples (from cuda/Dockerfile.samples)\n ↓\ncudastack:minimal (from cudastack/Dockerfile)\n ├── cuDNN\n ├── NCCL\n └── (all installed in ONE RUN)\n ↓\nyour-app\n```\n\n## Variants\n\n| Variant | Includes | Use Case |\n|---------|----------|----------|\n| `cudastack:minimal` | cuDNN + NCCL | Training, inference |\n| `cudastack:standard` | + TensorRT | Optimized inference |\n| `cudastack:full` | + CUTLASS + GDRCopy | Research, development |\n\n## Usage\n\n### Build with jetson-containers:\n\n```bash\n# Build minimal stack\n./build.sh cudastack:minimal\n\n# Build standard (with TensorRT)\n./build.sh cudastack:standard\n\n# Build full (with CUTLASS)\n./build.sh cudastack:full\n\n# Use as base for your app\n./build.sh --base cudastack:minimal pytorch\n```\n\n### The cudastack Dockerfile does ONE RUN:\n\n```dockerfile\n# Copy all install/build scripts\nCOPY install/*.sh /tmp/cudastack/install/\nCOPY build/*.sh /tmp/cudastack/build/\n\n# ONE RUN command calls all scripts conditionally\nRUN set -ex && \\\n if [ \"$WITH_CUDNN\" = \"1\" ]; then \\\n ./install/install_cudnn.sh; \\\n fi && \\\n if [ \"$WITH_NCCL\" = \"1\" ]; then \\\n ./install/install_nccl.sh; \\\n fi && \\\n if [ \"$WITH_TENSORRT\" = \"1\" ]; then \\\n ./install/install_tensorrt.sh; \\\n fi && \\\n # ... more components ...\n rm -rf /tmp/cudastack\n```\n\n**Key**: All library installations happen in this ONE RUN = ONE Docker layer!\n\n## Setup\n\nThe install scripts need to be populated from existing packages:\n\n```bash\ncd packages/cuda/cudastack\n./setup-scripts.sh\n```\n\nThen review and complete the scripts in `install/` and `build/`.\n\n## What About pip_cache, samples, builtin?\n\nThese are **still available** from the `cuda` package:\n\n```bash\n# pip_cache is automatically included by cuda package\n./build.sh cuda:12.4\n# This creates both:\n# - cuda:12.4 (main CUDA toolkit)\n# - pip_cache:cu124 (pip caching)\n\n# CUDA samples work as before\n./build.sh cuda:12.4-samples\n\n# Builtin works as before (for JetPack 4-5)\n./build.sh cuda:11.4 # Uses Dockerfile.builtin when CUDA pre-installed\n```\n\nThe `cuda` package's `config.py` returns multiple packages:\n```python\n# In cuda/config.py\ndef cuda_package(version, url, ...):\n cuda = package.copy()\n cuda['name'] = f'cuda:{version}'\n \n cuda_pip = pip_cache(version, requires) # Creates pip_cache package\n \n return cuda, cuda_pip # Returns TWO packages!\n```\n\n**cudastack doesn't need to duplicate this** - it just depends on `cuda` being installed first.\n\n## Component Toggles\n\nControlled by `WITH_*` build args (set in config.py):\n\n- ✅ `WITH_CUDNN=1` - cuDNN (always on)\n- ✅ `WITH_NCCL=1` - NCCL (always on)\n- 🎛️ `WITH_TENSORRT` - TensorRT (on in standard/full)\n- 🎛️ `WITH_CUTLASS` - CUTLASS (on in full only)\n- 🎛️ `WITH_GDRCOPY` - GDRCopy (on for Tegra)\n- ⚙️ `WITH_CUDSS=0` - cuDSS (off by default)\n- ⚙️ `WITH_CUSPARSELT=0` - cuSPARSELt (off by default)\n- ⚙️ `WITH_CUTENSOR=0` - cuTENSOR (off by default)\n- ⚙️ `WITH_NVPL=0` - NVPL (off by default)\n- ⚙️ `WITH_NVSHMEM=0` - NVSHMEM (off by default)\n\n## How config.py Works\n\nFollowing the jetson-containers pattern:\n\n```python\ndef cuda_stack(name, with_tensorrt=False, with_cutlass=False, requires=None):\n \"\"\"Generate a cudastack package variant\"\"\"\n pkg = package.copy()\n pkg['name'] = name\n pkg['depends'] = ['cuda'] # Depends ON cuda, doesn't replace it\n \n pkg['build_args'] = {\n 'WITH_CUDNN': '1',\n 'WITH_TENSORRT': '1' if with_tensorrt else '0',\n 'WITH_NCCL': '1',\n # ... etc\n }\n \n return pkg\n\n# Define variants\npackage = [\n cuda_stack('cudastack:minimal', with_tensorrt=False),\n cuda_stack('cudastack:standard', with_tensorrt=True),\n cuda_stack('cudastack:full', with_tensorrt=True, with_cutlass=True),\n]\n```\n\n## Comparison with Individual Packages\n\n| Approach | Layers | Build Time | Complexity |\n|----------|--------|------------|------------|\n| **Individual packages** | 50+ | Longer | Simple per-package |\n| **cudastack (consolidated)** | 2-3 | Faster | One complex layer |\n\nBoth approaches work! Use `cudastack` when you need multiple libraries and want to avoid layer limits.\n\n## See Also\n\n- [cuda package](../cuda/) - CUDA toolkit, pip_cache, samples, builtin\n- [cudnn package](../cudnn/) - Individual cuDNN package\n- [tensorrt package](../tensorrt/) - Individual TensorRT package (has Dockerfile.deb and Dockerfile.tar)\n- [nccl package](../nccl/) - Individual NCCL package\n\n## Summary\n\n✅ **cudastack is a consolidation layer** on top of cuda \n✅ **pip_cache, samples, builtin still work** from the cuda package \n✅ **ONE RUN command** installs all libraries to avoid layer limits \n✅ **Follows jetson-containers patterns** with config.py and variants \n","packages/cuda/cutlass/README.md":"# cutlass\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cutlass:4.1.0`** | |\n| :-- | :-- |\n|    Aliases | `cutlass` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cutlass)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cutlass:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cutlass)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cutlass) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cutlass\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cuda/pycuda/README.md":"# pycuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 0.4GB)`<br>[`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-09-07, 5.0GB)`<br>[`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-08-29, 5.0GB)`<br>[`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 5.0GB)`<br>[`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 3.5GB)`<br>[`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2025-03-03, 2.3GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `0.4GB` |\n|   [`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-09-07` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycuda)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/pycuda:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/pycuda:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dobjects/partpacker/README.md":"# partpacker\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`partpacker:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `partpacker` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/PartPacker |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag partpacker)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host partpacker:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag partpacker)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag partpacker) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build partpacker\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dobjects/sparc3d/README.md":"# sparc3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparc3d:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparc3d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/lizhihao6/Sparc3D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparc3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparc3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparc3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparc3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparc3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/glomap/README.md":"# glomap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`glomap:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `glomap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag glomap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host glomap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag glomap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag glomap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build glomap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/hloc/README.md":"# hloc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`hloc:1.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n| **`hloc:1.5`** | |\n| :-- | :-- |\n|    Aliases | `hloc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag hloc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host hloc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag hloc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag hloc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build hloc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/kaolin/README.md":"# kaolin\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`kaolin:0.18.0`** | |\n| :-- | :-- |\n|    Aliases | `kaolin` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVIDIAGameWorks/kaolin |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag kaolin)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host kaolin:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag kaolin)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag kaolin) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build kaolin\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/meshlab/README.md":"# meshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`meshlab:MeshLab-2023.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n| **`meshlab:MeshLab-2025.03`** | |\n| :-- | :-- |\n|    Aliases | `meshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag meshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host meshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag meshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag meshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build meshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/nerfacc/README.md":"# nerfacc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfacc:0.5.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n| **`nerfacc:0.5.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfacc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfacc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfacc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfacc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfacc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfacc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/open3d/README.md":"# open3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`open3d:1.19.0`** | |\n| :-- | :-- |\n|    Aliases | `open3d` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag open3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host open3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag open3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag open3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build open3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pixsfm/README.md":"# pixsfm\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pixsfm:1.0`** | |\n| :-- | :-- |\n|    Aliases | `pixsfm` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pixsfm)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pixsfm:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pixsfm)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pixsfm) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pixsfm\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/polyscope/README.md":"# polyscope\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`polyscope:2.5.0`** | |\n| :-- | :-- |\n|    Aliases | `polyscope` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nmwsharp/polyscope-py.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag polyscope)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host polyscope:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag polyscope)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag polyscope) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build polyscope\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pyceres/README.md":"# pyceres\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pyceres:2.5`** | |\n| :-- | :-- |\n|    Aliases | `pyceres` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`ninja`](/packages/build/ninja) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/pyceres.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pyceres)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pyceres:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pyceres)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pyceres) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pyceres\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pycolmap/README.md":"# pycolmap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycolmap:3.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n| **`pycolmap:3.13`** | |\n| :-- | :-- |\n|    Aliases | `pycolmap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycolmap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pycolmap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycolmap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycolmap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycolmap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/pymeshlab/README.md":"# pymeshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pymeshlab:2023.12.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2023.12.post3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2025.6.23.dev0`** | |\n| :-- | :-- |\n|    Aliases | `pymeshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pymeshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pymeshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pymeshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pymeshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pymeshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/3dvision/tinycudann/README.md":"# tinycudann\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`tinycudann:1.7`** | |\n| :-- | :-- |\n|    Aliases | `tinycudann` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/tiny-cuda-nn |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag tinycudann)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host tinycudann:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag tinycudann)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag tinycudann) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build tinycudann\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/3dgrut/README.md":"# 3dgrut\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3dgrut:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `3dgrut` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cupy`](/packages/numeric/cupy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`kaolin`](/packages/3d/3dvision/kaolin) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nv-tlabs/3dgrut |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3dgrut)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 3dgrut:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3dgrut)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3dgrut) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3dgrut\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/4k4d/README.md":"# 4k4d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`4k4d:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `4k4d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) [`easyvolcap`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/4k4D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 4k4d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 4k4d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 4k4d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 4k4d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 4k4d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/easyvolcap/README.md":"# easyvolcap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`easyvolcap:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `easyvolcap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/EasyVolcap |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag easyvolcap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host easyvolcap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag easyvolcap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag easyvolcap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build easyvolcap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/fast_gauss/README.md":"# fast_gauss\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fast_gauss:1.0.0`** | |\n| :-- | :-- |\n|    Aliases | `fast_gauss` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/dendenxu/fast-gaussian-rasterization |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fast_gauss)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host fast_gauss:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fast_gauss)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fast_gauss) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fast_gauss\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/gsplat/README.md":"# gsplat\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`gsplat:1.5.3`** | |\n| :-- | :-- |\n|    Aliases | `gsplat` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/gsplat |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag gsplat)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host gsplat:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag gsplat)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag gsplat) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build gsplat\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/gaussian_splatting/nerfview/README.md":"# nerfview\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfview:0.1.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfview` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`tinycudann`](/packages/3d/3dvision/tinycudann) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://pypi.org/project/nerfview/ |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfview)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfview:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfview)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfview) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfview\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/nerf/fruitnerf/README.md":"# fruitnerf\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fruitnerf:1.0`** | |\n| :-- | :-- |\n|    Aliases | `fruitnerf` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio`](/packages/3d/nerf/nerfstudio) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/meyerls/FruitNeRF |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/fruitnerf:r36.3.0`](https://hub.docker.com/r/dustynv/fruitnerf/tags) | `2024-09-02` | `arm64` | `14.6GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fruitnerf)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/fruitnerf:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/fruitnerf:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fruitnerf)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fruitnerf) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fruitnerf\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/3d/nerf/nerfstudio/README.md":"# nerfstudio\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfstudio:1.1.7`** | |\n| :-- | :-- |\n|    Aliases | `nerfstudio` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfstudio |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/nerfstudio:r36.3.0`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2024-09-02` | `arm64` | `7.7GB` |\n|   [`dustynv/nerfstudio:r36.4.0-cu128`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2025-02-11` | `arm64` | `7.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfstudio)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/nerfstudio:r36.4.0-cu128\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/nerfstudio:r36.4.0-cu128\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfstudio)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfstudio) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfstudio\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/cv-cuda/README.md":"# cv-cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cv-cuda:0.15`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15` `cv-cuda` `cvcuda` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg:git`](/packages/multimedia/ffmpeg) [`pyav`](/packages/multimedia/pyav) [`pycuda`](/packages/cuda/pycuda) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cv-cuda:0.15-cpp`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15-cpp` `cv-cuda:cpp` `cvcuda:cpp` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) |\n|    Dependants | [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cv-cuda)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cv-cuda:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cv-cuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cv-cuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cv-cuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/deepstream/README.md":"# deepstream\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`deepstream`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`cmake`](/packages/build/cmake/cmake_pip) [`numpy`](/packages/numeric/numpy) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`gstreamer`](/packages/multimedia/gstreamer) [`tritonserver`](/packages/ml/tritonserver) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-06, 2.3GB)`<br>[`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-22, 6.8GB)`<br>[`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 6.8GB)`<br>[`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-06, 6.7GB)`<br>[`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 9.8GB)` |\n|    Notes | https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_Quickstart.html |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-06` | `arm64` | `2.3GB` |\n|   [`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-22` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-06` | `arm64` | `6.7GB` |\n|   [`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `9.8GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag deepstream)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/deepstream:r32.7.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/deepstream:r32.7.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag deepstream)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag deepstream) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build deepstream\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/3d_diffusion_policy/README.md":"# 3d_diffusion_policy\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n\n3D Diffusion Policy: Generalizable Visuomotor Policy Learning (https://github.com/YanjieZe/3D-Diffusion-Policy)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3d_diffusion_policy`** | |\n| :-- | :-- |\n|    Aliases | `3d-diffusion-policy` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`mujoco`](/packages/sim/mujoco) [`numba`](/packages/numeric/numba) [`diffusers`](/packages/diffusion/diffusers) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) `(2024-09-30, 7.0GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) | `2024-09-30` | `arm64` | `7.0GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3d_diffusion_policy)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/3d_diffusion_policy:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/3d_diffusion_policy:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3d_diffusion_policy)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3d_diffusion_policy) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3d_diffusion_policy\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/ai-toolkit/README.md":"# ai-toolkit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\ndocs.md\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`ai-toolkit`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ai-toolkit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ai-toolkit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ai-toolkit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ai-toolkit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ai-toolkit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/cache_edit/README.md":"# cache_edit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cache_dit:0.2.0`** | |\n| :-- | :-- |\n|    Aliases | `cache_dit` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/vipshop/cache-dit |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cache_edit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cache_edit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cache_edit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cache_edit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cache_edit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","packages/cv/diffusion/comfyui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fcomfyui%2FREADME.md -->","packages/cv/diffusion/diffusers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusers%2FREADME.md -->","packages/cv/diffusion/diffusion_policy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusion_policy%2FREADME.md -->","packages/cv/diffusion/fastgen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Ffastgen%2FREADME.md -->","packages/cv/diffusion/framepack/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fframepack%2FREADME.md -->","packages/cv/diffusion/sdnext/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fsdnext%2FREADME.md -->","packages/cv/diffusion/self-forcing/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fself-forcing%2FREADME.md -->","packages/cv/diffusion/stable-diffusion-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fstable-diffusion-webui%2FREADME.md -->","packages/cv/holoscan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fholoscan%2FREADME.md -->","packages/cv/jetson-inference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fjetson-inference%2FREADME.md -->","packages/cv/opencv/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fopencv%2FREADME.md -->","packages/cv/vpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fvpi%2FREADME.md -->","packages/hw/canable/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanable%2FREADME.md -->","packages/hw/canbus/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanbus%2FREADME.md -->","packages/hw/cangaroo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcangaroo%2FREADME.md -->","packages/hw/jetcam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjetcam%2FREADME.md -->","packages/hw/jupyter_clickable_image_widget/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjupyter_clickable_image_widget%2FREADME.md -->","packages/hw/librealsense/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Flibrealsense%2FREADME.md -->","packages/hw/oled/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Foled%2FREADME.md -->","packages/hw/pl2303/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fpl2303%2FREADME.md -->","packages/hw/zed/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fzed%2FREADME.md -->","packages/llm/awq/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fawq%2FREADME.md -->","packages/llm/bitsandbytes/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fbitsandbytes%2FREADME.md -->","packages/llm/deepspeed/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2FREADME.md -->","packages/llm/deepspeed/deepspeed-kernels/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2Fdeepspeed-kernels%2FREADME.md -->","packages/llm/dynamo/dynamo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fdynamo%2FREADME.md -->","packages/llm/dynamo/kai-scheduler/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fkai-scheduler%2FREADME.md -->","packages/llm/dynamo/mooncake/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fmooncake%2FREADME.md -->","packages/llm/dynamo/nixl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fnixl%2FREADME.md -->","packages/llm/exllama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fexllama%2FREADME.md -->","packages/llm/gptqmodel/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fgptqmodel%2FREADME.md -->","packages/llm/huggingface_hub/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhuggingface_hub%2FREADME.md -->","packages/llm/hymba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhymba%2FREADME.md -->","packages/llm/ktransformers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fktransformers%2FREADME.md -->","packages/llm/llama-factory/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama-factory%2FREADME.md -->","packages/llm/llama_cpp/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama_cpp%2FREADME.md -->","packages/llm/llamaspeak/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllamaspeak%2FREADME.md -->","packages/llm/lobe_chat/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flobe_chat%2FREADME.md -->","packages/llm/local_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flocal_llm%2FREADME.md -->","packages/llm/minference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fminference%2FREADME.md -->","packages/llm/mistral/mistral_rs/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmistral%2Fmistral_rs%2FREADME.md -->","packages/llm/mlc/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmlc%2FREADME.md -->","packages/llm/nano_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnano_llm%2FREADME.md -->","packages/llm/nemo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnemo%2FREADME.md -->","packages/llm/ollama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Follama%2FREADME.md -->","packages/llm/open-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopen-webui%2FREADME.md -->","packages/llm/openai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopenai%2FREADME.md -->","packages/llm/optimum/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Foptimum%2FREADME.md -->","packages/llm/sglang/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2FREADME.md -->","packages/llm/sglang/genai-bench/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fgenai-bench%2FREADME.md -->","packages/llm/sglang/sgl-kernel/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fsgl-kernel%2FREADME.md -->","packages/llm/sudonim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsudonim%2FREADME.md -->","packages/llm/tensorrt_optimizer/nvidia-modelopt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Fnvidia-modelopt%2FREADME.md -->","packages/llm/tensorrt_optimizer/tensorrt_llm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Ftensorrt_llm%2FREADME.md -->","packages/llm/text-generation-inference/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-inference%2FREADME.md -->","packages/llm/text-generation-webui/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-webui%2FREADME.md -->","packages/llm/transformers/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftransformers%2FREADME.md -->","packages/llm/vllm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fvllm%2FREADME.md -->","packages/llm/xgrammar/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fxgrammar%2FREADME.md -->","packages/ml/aim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Faim%2FREADME.md -->","packages/ml/apache/tvm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fapache%2Ftvm%2FREADME.md -->","packages/ml/ctranslate2/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fctranslate2%2FREADME.md -->","packages/ml/dli/dli-nano-ai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fdli%2Fdli-nano-ai%2FREADME.md -->","packages/ml/jax/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fjax%2FREADME.md -->","packages/ml/kans/kat/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fkat%2FREADME.md -->","packages/ml/kans/pykan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fpykan%2FREADME.md -->","packages/ml/l4t/l4t-diffusion/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-diffusion%2FREADME.md -->","packages/ml/l4t/l4t-dynamo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-dynamo%2FREADME.md -->","packages/ml/l4t/l4t-ml/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-ml%2FREADME.md -->","packages/ml/l4t/l4t-pytorch/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-pytorch%2FREADME.md -->","packages/ml/l4t/l4t-tensorflow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-tensorflow%2FREADME.md -->","packages/ml/l4t/l4t-text-generation/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-text-generation%2FREADME.md -->","packages/ml/mamba/causalconv1d/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcausalconv1d%2FREADME.md -->","packages/ml/mamba/cobra/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcobra%2FREADME.md -->","packages/ml/mamba/dimba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fdimba%2FREADME.md -->","packages/ml/mamba/mamba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmamba%2FREADME.md -->","packages/ml/mamba/mambavision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmambavision%2FREADME.md -->","packages/ml/mamba/videomambasuite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fvideomambasuite%2FREADME.md -->","packages/ml/mamba/zigma/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fzigma%2FREADME.md -->","packages/ml/numeric/arrow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Farrow%2FREADME.md -->","packages/ml/numeric/cupy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fcupy%2FREADME.md -->","packages/ml/numeric/numba/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumba%2FREADME.md -->","packages/ml/numeric/numpy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumpy%2FREADME.md -->","packages/ml/numeric/warp/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fwarp%2FREADME.md -->","packages/ml/onnx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnx%2FREADME.md -->","packages/ml/onnxruntime/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime%2FREADME.md -->","packages/ml/onnxruntime_genai/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime_genai%2FREADME.md -->","packages/ml/pytorch/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2FREADME.md -->","packages/ml/pytorch/apex/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Fapex%2FREADME.md -->","packages/ml/pytorch/torch2trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch2trt%2FREADME.md -->","packages/ml/pytorch/torch3d/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch3d%2FREADME.md -->","packages/ml/pytorch/torch_tensorrt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch_tensorrt%2FREADME.md -->","packages/ml/pytorch/torchao/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchao%2FREADME.md -->","packages/ml/pytorch/torchaudio/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchaudio%2FREADME.md -->","packages/ml/pytorch/torchsaver/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsaver%2FREADME.md -->","packages/ml/pytorch/torchsde/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsde%2FREADME.md -->","packages/ml/pytorch/torchtext/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchtext%2FREADME.md -->","packages/ml/pytorch/torchvision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchvision%2FREADME.md -->","packages/ml/rapids/cudf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcudf%2FREADME.md -->","packages/ml/rapids/cuml/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcuml%2FREADME.md -->","packages/ml/rapids/raft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fraft%2FREADME.md -->","packages/ml/tensorflow/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2FREADME.md -->","packages/ml/tensorflow/graphics/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Fgraphics%2FREADME.md -->","packages/ml/tensorflow/tensorboard/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftensorboard%2FREADME.md -->","packages/ml/tensorflow/text/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftext%2FREADME.md -->","packages/ml/transformer-engine/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftransformer-engine%2FREADME.md -->","packages/ml/triton/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftriton%2FREADME.md -->","packages/ml/tritonserver/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftritonserver%2FREADME.md -->","packages/ml/xlstm/mlstm_kernels/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fmlstm_kernels%2FREADME.md -->","packages/ml/xlstm/pltsm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fpltsm%2FREADME.md -->","packages/ml/xlstm/xlstm/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fxlstm%2FREADME.md -->","packages/multimedia/decord/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fdecord%2FREADME.md -->","packages/multimedia/ffmpeg/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fffmpeg%2FREADME.md -->","packages/multimedia/gstreamer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fgstreamer%2FREADME.md -->","packages/multimedia/jetson-utils/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fjetson-utils%2FREADME.md -->","packages/multimedia/libcom/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Flibcom%2FREADME.md -->","packages/multimedia/opengl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fopengl%2FREADME.md -->","packages/multimedia/pyav/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fpyav%2FREADME.md -->","packages/multimedia/sound-utils/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fsound-utils%2FREADME.md -->","packages/multimedia/video-codec-sdk/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvideo-codec-sdk%2FREADME.md -->","packages/multimedia/vulkan/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvulkan%2FREADME.md -->","packages/net/devpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fdevpi%2FREADME.md -->","packages/net/pypi.dev/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fpypi.dev%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-policy/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-policy%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-predict2/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-predict2%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-reason1/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-reason1%2FREADME.md -->","packages/physicalAI/cosmos/cosmos-transfer1/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-transfer1%2FREADME.md -->","packages/physicalAI/cosmos/cosmos_diffusion_renderer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2FREADME.md -->","packages/physicalAI/cosmos/cosmos_diffusion_renderer/nvdiffrast/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2Fnvdiffrast%2FREADME.md -->","packages/physicalAI/isaac-ros/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fisaac-ros%2FREADME.md -->","packages/physicalAI/lerobot/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Flerobot%2FREADME.md -->","packages/physicalAI/opendronemap/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopendronemap%2FREADME.md -->","packages/physicalAI/openpi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopenpi%2FREADME.md -->","packages/physicalAI/protomotions/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fprotomotions%2FREADME.md -->","packages/physicalAI/ros/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fros%2FREADME.md -->","packages/physicalAI/sim/genesis/Readme.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FReadme.md -->","packages/physicalAI/sim/genesis/quadrants/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fquadrants%2FREADME.md -->","packages/physicalAI/sim/genesis/splashSurf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FsplashSurf%2FREADME.md -->","packages/physicalAI/sim/genesis/vtk/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fvtk%2FREADME.md -->","packages/physicalAI/sim/habitat-sim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fhabitat-sim%2FREADME.md -->","packages/physicalAI/sim/isaac-sim/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2FREADME.md -->","packages/physicalAI/sim/isaac-sim/isaac-lab/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2Fisaac-lab%2FREADME.md -->","packages/physicalAI/sim/mimicgen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmimicgen%2FREADME.md -->","packages/physicalAI/sim/mujoco/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmujoco%2FREADME.md -->","packages/physicalAI/sim/newton/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fnewton%2FREADME.md -->","packages/physicalAI/sim/physx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fphysx%2FREADME.md -->","packages/physicalAI/sim/robogen/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobogen%2FREADME.md -->","packages/physicalAI/sim/robomimic/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobomimic%2FREADME.md -->","packages/physicalAI/sim/robosuite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobosuite%2FREADME.md -->","packages/physicalAI/sim/sapien/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fsapien%2FREADME.md -->","packages/physicalAI/vla/crossformer/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fcrossformer%2FREADME.md -->","packages/physicalAI/vla/isaac-gr00t/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fisaac-gr00t%2FREADME.md -->","packages/physicalAI/vla/octo/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Focto%2FREADME.md -->","packages/physicalAI/vla/openvla/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fopenvla%2FREADME.md -->","packages/physicalAI/vla/robopoint/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Frobopoint%2FREADME.md -->","packages/rag/graphiti/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fgraphiti%2FREADME.md -->","packages/rag/jetson-copilot/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fjetson-copilot%2FREADME.md -->","packages/rag/langchain/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Flangchain%2FREADME.md -->","packages/rag/llama-index/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fllama-index%2FREADME.md -->","packages/rag/memvid/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fmemvid%2FREADME.md -->","packages/smart-home/dependencies/ciso8601/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fciso8601%2FREADME.md -->","packages/smart-home/dependencies/psutil-home-assistant/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fpsutil-home-assistant%2FREADME.md -->","packages/smart-home/homeassistant-base/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-base%2FREADME.md -->","packages/smart-home/homeassistant-core/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-core%2FREADME.md -->","packages/smart-home/wyoming/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2FREADME.md -->","packages/smart-home/wyoming/wyoming-assist-microphone/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-assist-microphone%2FREADME.md -->","packages/smart-home/wyoming/wyoming-openwakeword/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-openwakeword%2FREADME.md -->","packages/smart-home/wyoming/wyoming-piper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-piper%2FREADME.md -->","packages/smart-home/wyoming/wyoming-whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-whisper%2FREADME.md -->","packages/speech/audiocraft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Faudiocraft%2FREADME.md -->","packages/speech/chatterbox-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fchatterbox-tts%2FREADME.md -->","packages/speech/espeak/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fespeak%2FREADME.md -->","packages/speech/faster-whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Ffaster-whisper%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-fastapi/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-fastapi%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-hf/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-hf%2FREADME.md -->","packages/speech/kokoro-tts/kokoro-tts-onnx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-onnx%2FREADME.md -->","packages/speech/piper1-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fpiper1-tts%2FREADME.md -->","packages/speech/riva-client/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Friva-client%2FREADME.md -->","packages/speech/spark-tts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspark-tts%2FREADME.md -->","packages/speech/speaches/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeaches%2FREADME.md -->","packages/speech/speech-dispatcher/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeech-dispatcher%2FREADME.md -->","packages/speech/voice-pro/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoice-pro%2FREADME.md -->","packages/speech/voicecraft/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoicecraft%2FREADME.md -->","packages/speech/whisper/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper%2FREADME.md -->","packages/speech/whisper_trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper_trt%2FREADME.md -->","packages/speech/whisperx/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisperx%2FREADME.md -->","packages/speech/xtts/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fxtts%2FREADME.md -->","packages/vectordb/faiss/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss%2FREADME.md -->","packages/vectordb/faiss_lite/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss_lite%2FREADME.md -->","packages/vectordb/nanodb/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Fnanodb%2FREADME.md -->","packages/vit/clip_trt/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fclip_trt%2FREADME.md -->","packages/vit/efficientvit/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fefficientvit%2FREADME.md -->","packages/vit/nanoowl/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanoowl%2FREADME.md -->","packages/vit/nanosam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanosam%2FREADME.md -->","packages/vit/sam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsam%2FREADME.md -->","packages/vit/sapiens/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsapiens%2FREADME.md -->","packages/vit/tam/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Ftam%2FREADME.md -->","packages/vlm/lita/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Flita%2FREADME.md -->","packages/vlm/llama-vision/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllama-vision%2FREADME.md -->","packages/vlm/llava/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllava%2FREADME.md -->","packages/vlm/minigpt4/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fminigpt4%2FREADME.md -->","packages/vlm/prismatic/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fprismatic%2FREADME.md -->","packages/vlm/videollama/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvideollama%2FREADME.md -->","packages/vlm/vila-microservice/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila-microservice%2FREADME.md -->","packages/vlm/vila/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila%2FREADME.md -->","packages/vlm/xtuner/README.md":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fxtuner%2FREADME.md -->"},"items":[{"name":"AGENTS.md","path":"AGENTS.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/AGENTS.md","title":"AI Agent Protocol & Instructions","category":"root-instruction","format":"markdown","content":"# CLAUDE.md\n\nThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.\n\n## Overview\n\njetson-containers is a modular container build system for NVIDIA Jetson devices. It discovers packages, resolves their dependencies, chains their Dockerfiles together, and runs `docker build` to produce multi-stage images optimized for specific L4T/JetPack/CUDA versions.\n\n## Commands\n\n### Build & Run\n```bash\n# Build one or more packages (chains Dockerfiles in dependency order)\njetson-containers build <package> [<package2> ...]\n# or equivalently:\n./build.sh <package> [<package2> ...]\n\n# Find and run a compatible container image\njetson-containers run $(autotag <package>) [command]\n# or equivalently:\n./run.sh $(autotag <package>) [command]\n\n# Find the best matching image tag (local → DockerHub → build)\nautotag <package>\n\n# Useful build flags\njetson-containers build --simulate <package> # dry-run: print commands without building\njetson-containers build --multiple pytorch tensorflow # build each package as a separate image\njetson-containers build --list-packages # list all discovered packages\njetson-containers build --show-packages <package> # show resolved metadata for a package\njetson-containers build --base=xyz:latest <package> # start the build chain from an existing image\njetson-containers build --push=myuser <package> # push to a registry after building\n\n# Check for upstream version upgrades (--apply to write changes)\njetson-containers upgrade [<package> ...] [--apply]\n```\n\n### Code Style\n```bash\n# Install pre-commit hooks (one-time setup)\npip install -r requirements.txt\npre-commit install\n\n# Run linting/formatting checks manually\npre-commit run --all-files\n```\n\nPre-commit runs **Black** (formatter, line-length=88) and **Flake8** (linter) only on `jetson_containers/` and `test_precommit.py`. `packages/` Dockerfiles and `config.py` files are **not** linted by pre-commit. Note: `skip-string-normalization = true` in `pyproject.toml`.\n\n### Tests\n```bash\n# Run a package config unit test directly (uses stdlib unittest, no pytest needed)\npython3 tests/verify_triton_config.py\n```\n\n### Installation\n```bash\nbash install.sh # sets up venv, installs deps, symlinks CLI tools to /usr/local/bin\n```\n\n## Architecture\n\n### Package System\n\nEvery buildable unit is a **package**. Packages live under `packages/<category>/<name>/` and define metadata via one of (in priority order):\n\n1. **YAML header in Dockerfile** — metadata between `#---` markers at the top\n2. **`config.yaml` / `config.yml`** — static YAML config file\n3. **`config.json`** — meta-container (no Dockerfile, just dependency composition)\n4. **`config.py`** — dynamic config, returns a list of package dicts at build time\n\nKey metadata fields: `name`, `alias`, `depends` (build-order deps), `requires` (version constraints on L4T/CUDA/Python), `test`, `build_args`, `dockerfile`.\n\n### Build Pipeline\n\n1. **`jetson_containers/packages.py`** — scans `packages/` recursively, runs `config.py` scripts, resolves the dependency graph, filters by L4T version compatibility, and exposes a custom Python meta path finder so packages can `from packages.ml.pytorch.version import ...`\n2. **`jetson_containers/build.py`** — CLI entry point; parses args and dispatches\n3. **`jetson_containers/container.py`** — core build logic: stitches multiple package Dockerfiles into one combined `Dockerfile`, invokes `docker build`, optionally runs per-package tests, and optionally pushes to a registry\n4. **`jetson_containers/l4t_version.py`** — detects architecture (tegra-aarch64, aarch64, x86_64) and reads L4T/JetPack/CUDA/GPU-arch info from `/etc/nv_tegra_release`; all version variables (e.g. `CUDA_VERSION`, `PYTORCH_VERSION`) flow from here into build args\n\n### Dynamic Package Generation\n\n`config.py` files execute at build time and can return multiple package dicts to produce version variants. For example, PyTorch's `config.py` generates `pytorch:2.8`, `pytorch:2.8-all`, `pytorch:2.8-builder`, etc. Aliases (`torch` → `pytorch:2.8`) are resolved during package discovery.\n\nWhen a package's version selection logic is complex, put it in a separate `version.py` alongside `config.py`. This lets other packages cross-import it:\n```python\nfrom packages.ml.pytorch.version import PYTORCH_VERSION\n```\nThe `_PackagesFinder` meta path finder in `packages.py` resolves these imports against the `packages/` tree without requiring `__init__.py` files.\n\n### Dockerfile Conventions\n\nUse `uv pip install` (not plain `pip install`) inside Dockerfiles — `uv` is pre-installed in the base images and is significantly faster. Every Dockerfile must start with:\n```dockerfile\nARG BASE_IMAGE\nFROM ${BASE_IMAGE}\n```\n\n### Version Overrides\n\nCopy `.env.default` → `.env` to override `L4T_VERSION`, `CUDA_VERSION`, `LSB_RELEASE`, `CUDA_ARCH`, local PyPI/APT mirror URLs, and SCP/webhook credentials without modifying tracked files.\n\n### Creating a New Package\n\n1. Create `packages/<category>/<name>/Dockerfile` with a `#---` YAML header (or a `config.yaml`/`config.py`).\n2. Declare `depends` on upstream packages (e.g., `pytorch`, `cuda`).\n3. Optionally add `test.py` or `test.sh` for post-build validation.\n4. Build: `jetson-containers build <name>`\n\n## Security Features\n\n### Multi-Stage Builds\n\nThe build system is multi-stage aware. Packages can use `FROM ... AS <alias>` for intermediate stages (e.g. a compiler/builder stage) without those stages polluting the final image. The build system tracks stage boundaries and only injects BuildKit features (ccache env vars, `--device` mounts, ccache `--mount` flags) into the **final** stage (the `FROM ${BASE_IMAGE}` stage). Intermediate stages are left untouched.\n\nDeclare multi-stage packages in config metadata:\n```yaml\n#---\n# name: mypackage\n# multistage: true\n# depends: [cuda]\n#---\nARG BASE_IMAGE\nFROM gcc:12 AS builder\nRUN make install\n\nFROM ${BASE_IMAGE}\nCOPY --from=builder /usr/local /usr/local\nRUN ldconfig\n```\n\nThe final stage **must** use `ARG BASE_IMAGE` / `FROM ${BASE_IMAGE}` so it chains correctly in the dependency build pipeline.\n\n### Non-Root User\n\n**At build time** — pass identity via env vars; packages can create and switch to the declared user:\n```bash\nexport CONTAINER_USER=jetson\nexport CONTAINER_UID=1000\nexport CONTAINER_GID=1000\njetson-containers build mypackage\n```\n\n`CONTAINER_USER`, `CONTAINER_UID`, `CONTAINER_GID` are forwarded as `--build-arg` automatically. Packages consume them like:\n```dockerfile\nARG CONTAINER_USER=root\nARG CONTAINER_UID=0\nARG CONTAINER_GID=0\nRUN if [ \"$CONTAINER_UID\" != \"0\" ]; then \\\n groupadd -g $CONTAINER_GID $CONTAINER_USER && \\\n useradd -u $CONTAINER_UID -g $CONTAINER_GID -m $CONTAINER_USER; \\\n fi\nUSER $CONTAINER_USER\n```\n\nDeclare the supported user in package metadata (`run_user` is informational for tooling):\n```yaml\n# run_user: jetson\n```\n\n**At runtime** — set `DOCKER_USER` to run the container as a non-root identity:\n```bash\nDOCKER_USER=1000:1000 ./run.sh myimage\n# or by name (the user must exist in the image):\nDOCKER_USER=jetson ./run.sh myimage\n```\nX11 forwarding (`xhost`) is automatically updated to grant the declared user display access.\n\nFor GPU access, `run.sh` also discovers the numeric GIDs that own the NVIDIA/Tegra device nodes (`/dev/nvhost*`, `/dev/nvmap`, `/dev/nvgpu/*`, `/dev/dri/render*`) and adds them via `--group-add <gid>`. Without this, a non-root process can't open the GPU nodes and CUDA fails with error 801 (`cudaGetDeviceCount` \"operation not supported\"). Numeric GIDs are used because the host's `render` GID usually doesn't match the container's. See [`docs/run.md`](/docs/run.md#gpu-access-as-non-root-cuda-error-801).\n\n### Secrets\n\n**Build-time secrets** (never baked into image layers) — declare secrets in package metadata:\n```yaml\n#---\n# name: mypackage\n# secrets: [huggingface_token]\n#---\n```\n\nPoint each secret at a file on the host via `JETSON_SECRET_<NAME>`:\n```bash\nexport JETSON_SECRET_HUGGINGFACE_TOKEN=~/.hf_token\njetson-containers build mypackage\n```\n\nSecrets are mounted as `/run/secrets/<name>` inside `RUN` steps only (BuildKit `--secret`). Use them in the Dockerfile:\n```dockerfile\nRUN --mount=type=secret,id=huggingface_token \\\n HF_TOKEN=$(cat /run/secrets/huggingface_token) \\\n python3 download_model.py\n```\n\nIf the env var is missing, a warning is printed and the build continues without that secret.\n\n**Runtime secrets** — prefer file-backed tokens over env vars (values in env vars are visible in `docker inspect`):\n```bash\n# Preferred: mount token file into /run/secrets/\nHUGGINGFACE_TOKEN_FILE=~/.hf_token ./run.sh myimage\n\n# Fallback (token visible in docker inspect):\nHUGGINGFACE_TOKEN=hf_xxx ./run.sh myimage\n```\n\n`HUGGINGFACE_TOKEN_FILE` mounts the file read-only at `/run/secrets/huggingface_token` and sets `HF_TOKEN_FILE` inside the container.\n\n### Key Supporting Modules\n\n- `jetson_containers/tag.py` — image tagging conventions\n- `jetson_containers/docs.py` — auto-generates per-package docs from metadata\n- `jetson_containers/ci.py` — CI/CD helpers\n- `jetson_containers/network.py` / `webhook.py` — GitHub API integration and build notifications\n- `jetson_containers/logging.py` — colored terminal output used throughout\n- `jetson_containers/upgrade.py` — checks packages against upstream (PyPI/GitHub) for newer versions; powers `jetson-containers upgrade`\n- `jetson_containers/db.py` — syncs DockerHub/GitHub metadata for registry introspection\n","isInternal":false,"tokens":2402,"sizeBytes":9558},{"name":"CLAUDE.md","path":"CLAUDE.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/CLAUDE.md","title":"Claude Agent Guidelines & System Prompt","category":"claude-rule","format":"markdown","content":"@AGENTS.md\n","isInternal":false,"tokens":5,"sizeBytes":11},{"name":"jetson-pr.md","path":".claude/skills/jetson-pr.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/.claude/skills/jetson-pr.md","title":"Skills Skill","category":"anthropic-skill","format":"markdown","content":"# Create a jetson-containers Pull Request\n\nHelp the user create a well-formed PR that follows this repository's conventions.\n\n## Repository PR conventions\n\n- **Target branch**: PRs go to `dev`, not `master`. The CI workflow `pr-on-dev.yml` runs self-hosted Jetson builds against `dev`.\n- **Scope**: each PR should touch only the packages it claims to change. Keep unrelated cleanup out.\n- **Commit style**: short, lowercase, imperative subject. No period. Examples from git log: `fix`, `cuda 13.2 needs newer drivers to run in Orin`, `tensorrt compatible orin`, `Update ffmpeg version from 8.1 to 8.0.1`.\n- **No force-push** to shared branches.\n\n## Step-by-step workflow\n\n1. **Understand the changes**: run `git diff master...HEAD` and `git log master..HEAD --oneline` to see exactly what is changing. Read the changed files if needed to understand the intent.\n\n2. **Identify affected packages**: map every changed file back to its package directory (`packages/<category>/<name>/`). List them — this drives the PR title and body.\n\n3. **Stage and commit** (if there are uncommitted changes):\n - Stage only the files related to this PR.\n - Write a commit message following the style above. No trailing period.\n - Do not skip hooks (`--no-verify`).\n\n4. **Create or switch to a feature branch** if on `master`:\n ```\n git checkout -b <short-descriptive-name>\n ```\n Branch name: lowercase, hyphenated, describes the change (e.g. `upgrade-pytorch-2.9`, `add-sglang-0.4`, `fix-tensorrt-orin`).\n\n5. **Push and open the PR**:\n ```\n gh pr create --base dev --title \"<title>\" --body \"<body>\"\n ```\n\n6. **PR title format**: `<verb> <package(s)> <version or brief description>`\n - Version bumps: `upgrade pytorch to 2.9`\n - New packages: `add sglang 0.4`\n - Bug fixes: `fix tensorrt build on Orin`\n - Multi-package: list up to 3, then `and N others`\n\n7. **PR body template**:\n ```\n ## Changes\n - <package>: <what changed and why>\n\n ## Packages affected\n - `packages/<category>/<name>/`\n\n ## Test plan\n - [ ] `jetson-containers build <package>` on Orin (JP6)\n - [ ] Add any version-specific test steps here\n\n ## Notes\n <Any JetPack version constraints, driver requirements, or upstream release links>\n ```\n\n8. **Return the PR URL** when done.\n\n## What good PRs look like in this repo\n\n- Version upgrades include the new version entry in `config.py` **and** update `version.py` if it exists.\n- New packages include a `Dockerfile`, metadata (YAML header or `config.yaml`), and at minimum a stub `test.py` or `test.sh`.\n- Dependency bumps update `requires` constraints if the new version drops support for an older JetPack.\n- PRs do not mix version bumps with unrelated refactors.\n\n## Before opening the PR, verify\n\n- `git diff --stat dev...HEAD` shows only the intended files.\n- No `.env`, secrets, or build artifacts are staged.\n- The branch name and title are consistent.\n","isInternal":false,"tokens":751,"sizeBytes":2908},{"name":"upgrade-package.md","path":".claude/skills/upgrade-package.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/.claude/skills/upgrade-package.md","title":"Skills Skill","category":"anthropic-skill","format":"markdown","content":"# Upgrade Package Version\n\nHelp the user upgrade the version of one or more packages in this jetson-containers repository.\n\n## What the user will provide\n\nThe user will tell you the package name and the new version. If they only give you the package name, ask for the target version before proceeding.\n\n## How version definitions work\n\nEach package under `packages/<category>/<name>/` defines its version in one of these ways:\n\n1. **`version.py`** — a standalone file that exports a version constant, e.g. `PYTORCH_VERSION = Version('2.8')`. This is the canonical version used by `config.py`.\n2. **`config.py`** — calls a helper function (like `pytorch_pip(version, requires=...)`) with explicit version strings in a `package = [...]` list. Adding a new version = adding a new line to this list.\n3. **`config.yaml` / YAML header in Dockerfile** — static `build_args` with version pinned directly.\n4. **`Dockerfile` `ARG` lines** — `ARG SOME_VERSION=x.y.z`.\n\n## Step-by-step workflow\n\n1. **Locate the package directory**: search under `packages/` for the package name.\n2. **Read all config files** in that directory (`config.py`, `config.yaml`, `version.py`, `Dockerfile`, `Dockerfile.*`).\n3. **Identify where the version is defined** (see above). Show the user the current version(s) before making changes.\n4. **Apply the upgrade**:\n - If there is a `version.py`, update the version constant there first; `config.py` will pick it up automatically.\n - If `config.py` uses a list of `pytorch_pip(version, ...)` calls, **add** a new entry for the new version (do not remove old ones — the repo keeps historical versions for different JetPack targets). Set the new version as the latest by updating the `PYTORCH_VERSION`-like constant in `version.py`.\n - If the version is in a YAML `build_args` or a Dockerfile `ARG`, do a targeted edit of that value.\n5. **Check for cascading dependencies**: grep for the old version string across other config files in the same package directory. Common locations: `Dockerfile` `ARG` defaults, `build_args` dicts, `requires` strings, download URLs, wheel filenames.\n6. **Update download artifacts if needed**: if the Dockerfile fetches a wheel, tarball, or pre-built binary, update the URL/filename to match the new version. Check the upstream source (PyPI, NVIDIA, GitHub releases) if the user hasn't provided a URL.\n7. **Show a diff summary** of every file you changed before finishing.\n\n## Things to preserve\n\n- Do **not** delete existing version entries in `config.py` lists — old JetPack targets still need them.\n- Keep the `requires` constraints on old entries unchanged.\n- Do not change `alias` lists for older versions.\n- Follow the existing code style (no reformatting unrelated code, no added comments).\n\n## When to ask vs. proceed\n\n- Ask for the target version if not given.\n- Ask about JetPack `requires` constraint only if the new version introduces a minimum JetPack requirement that differs from the previous latest entry.\n- Proceed with everything else silently and show the diff at the end.\n","isInternal":false,"tokens":717,"sizeBytes":3058},{"name":"README.md","path":"packages/attention/block-sparse-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/block-sparse-attention/README.md","title":"block-sparse-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# block-sparse-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`block_sparse_attn:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `block_sparse_attn` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`xattention:0.0.1`](/packages/attention/xattention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag block-sparse-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host block-sparse-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag block-sparse-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag block-sparse-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build block-sparse-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1021,"sizeBytes":3653},{"name":"README.md","path":"packages/attention/flash-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/flash-attention/README.md","title":"flash-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# flash-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flash-attention:2.5.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.6.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.4.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.0.post2`** | |\n| :-- | :-- |\n|    Aliases | `flash-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":4368,"sizeBytes":12779},{"name":"README.md","path":"packages/attention/flashinfer/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/flashinfer/README.md","title":"flashinfer Documentation","category":"plugin-manifest","format":"markdown","content":"# flash-infer\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flashinfer:0.2.1.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:0.8.4`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.6.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`mlc:0.20.0`](/packages/llm/mlc) [`sglang:0.4.6`](/packages/llm/sglang) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.8`** | |\n| :-- | :-- |\n|    Aliases | `flashinfer` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`llama-factory`](/packages/llm/llama-factory) [`local_llm`](/packages/llm/local_llm) [`mlc:0.21.0`](/packages/llm/mlc) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-infer)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-infer:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-infer)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-infer) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-infer\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":2975,"sizeBytes":8485},{"name":"README.md","path":"packages/attention/flexprefill/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/flexprefill/README.md","title":"flexprefill Documentation","category":"plugin-manifest","format":"markdown","content":"# flexprefill\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flexprefill:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `flexprefill` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`llama-factory`](/packages/llm/llama-factory) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`sdnext`](/packages/diffusion/sdnext) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flexprefill)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flexprefill:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flexprefill)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flexprefill) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flexprefill\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1326,"sizeBytes":4270},{"name":"README.md","path":"packages/attention/huggingface_kernels/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/huggingface_kernels/README.md","title":"huggingface_kernels Documentation","category":"plugin-manifest","format":"markdown","content":"# huggingface_kernels\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`huggingface_kernels:0.7.0`** | |\n| :-- | :-- |\n|    Aliases | `huggingface_kernels` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`huggingface_hub`](/packages/llm/huggingface_hub) [`torchvision`](/packages/pytorch/torchvision) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag huggingface_kernels)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host huggingface_kernels:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag huggingface_kernels)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag huggingface_kernels) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build huggingface_kernels\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":968,"sizeBytes":3467},{"name":"README.md","path":"packages/attention/log-linear-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/log-linear-attention/README.md","title":"log-linear-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# log-linear-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`log-linear-attention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `log-linear-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`ninja`](/packages/build/ninja) [`torchaudio`](/packages/pytorch/torchaudio) [`causalconv1d`](/packages/ml/mamba/causalconv1d) [`mamba`](/packages/ml/mamba/mamba) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag log-linear-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host log-linear-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag log-linear-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag log-linear-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build log-linear-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":976,"sizeBytes":3502},{"name":"README.md","path":"packages/attention/ParaAttention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/ParaAttention/README.md","title":"ParaAttention Documentation","category":"plugin-manifest","format":"markdown","content":"# ParaAttention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nhttps://github.com/chengzeyi/ParaAttention\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`paraattention:0.4.0`** | |\n| :-- | :-- |\n|    Aliases | `paraattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ParaAttention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ParaAttention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ParaAttention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ParaAttention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ParaAttention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1126,"sizeBytes":3990},{"name":"README.md","path":"packages/attention/radial-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/radial-attention/README.md","title":"radial-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# radial-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`radial-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `radial-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`flashinfer`](/packages/attention/flash-infer) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag radial-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host radial-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag radial-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag radial-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build radial-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":929,"sizeBytes":3354},{"name":"README.md","path":"packages/attention/sage-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/sage-attention/README.md","title":"sage-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# sage-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sage-attention:3.0.0`** | |\n| :-- | :-- |\n|    Aliases | `sage-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`framepack`](/packages/diffusion/framepack) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sage-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sage-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sage-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sage-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sage-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1135,"sizeBytes":4000},{"name":"README.md","path":"packages/attention/sparge-attention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/sparge-attention/README.md","title":"sparge-attention Documentation","category":"plugin-manifest","format":"markdown","content":"# sparge-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparge-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparge-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/thu-ml/SpargeAttn.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparge-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparge-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparge-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparge-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparge-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1154,"sizeBytes":4042},{"name":"README.md","path":"packages/attention/xattention/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/xattention/README.md","title":"xattention Documentation","category":"plugin-manifest","format":"markdown","content":"# xattention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xattention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `xattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`block_sparse_attn`](/packages/attention/block-sparse-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xattention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host xattention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xattention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xattention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xattention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1056,"sizeBytes":3786},{"name":"README.md","path":"packages/attention/xformers/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/attention/xformers/README.md","title":"xformers Documentation","category":"plugin-manifest","format":"markdown","content":"# xformers\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xformers:0.0.32`** | |\n| :-- | :-- |\n|    Aliases | `xformers` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dependants | [`apex:0.1`](/packages/pytorch/apex) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n| **`xformers:0.0.33`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/xformers:r35.2.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-06` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.3.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-14` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.4.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2024-01-09` | `arm64` | `5.9GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xformers)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/xformers:r35.4.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/xformers:r35.4.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xformers)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xformers) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xformers\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":3796,"sizeBytes":10790},{"name":"README.md","path":"packages/code/jupyterlab/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/code/jupyterlab/README.md","title":"jupyterlab Documentation","category":"plugin-manifest","format":"markdown","content":"# jupyterlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nRunning the `jupyerlab` container will automatically start a JupyterLab server in the background on port 8888, with default login password `nvidia`. The JupyterLab server logs will be saved to `/data/logs/jupyter.log` if you need to inspect them (this location is automatically mounted under your `jetson-containers/data` directory)\n\nTo change the default settings, you can set the `$JUPYTER_ROOT`, `$JUPYTER_PORT`, `$JUPYTER_PASSWORD`, and `$JUPYTER_LOG` environment variables when starting the container like so:\n\n```bash\njetson-containers run \\\n --env JUPYTER_ROOT=/home/user \\\n --env JUPYTER_PORT=8000 \\\n --env JUPYTER_PASSWORD=password \\\n --env JUPYTER_LOGS=/dev/null \\\n $(autotag jupyterlab)\n```\n\nThe [`/start_jupyter`](./start_jupyter) script is the default CMD that the container runs when it starts - however, if you don't want the JupyterLab server started by default, you can either add a different CMD in your own Dockerfile, or override it at startup:\n\n```bash\n# skip straight to the terminal instead of starting JupyterLab first\njetson-containers run /bin/bash\n```\n\nYou can then still manually run the [`/start_jupyter`](./start_jupyter) script later when desired.\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`jupyterlab:latest`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:latest-myst`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab:myst` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:latest`](/packages/code/jupyterlab) |\n|    Dependants | [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0-myst`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/jupyterlab:r32.7.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.7GB` |\n|   [`dustynv/jupyterlab:r35.2.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-12-06` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r35.3.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `5.4GB` |\n|   [`dustynv/jupyterlab:r35.4.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-10-07` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r36.2.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.6GB` |\n|   [`dustynv/jupyterlab:r36.4.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-10` | `arm64` | `5.8GB` |\n|   [`dustynv/jupyterlab:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-03` | `arm64` | `5.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag jupyterlab)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/jupyterlab:r36.4.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/jupyterlab:r36.4.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag jupyterlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag jupyterlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build jupyterlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":3020,"sizeBytes":9603},{"name":"README.md","path":"packages/code/vscode/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/code/vscode/README.md","title":"vscode Documentation","category":"plugin-manifest","format":"markdown","content":"# vscode\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`vscode`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:cuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`tensorrt`](/packages/cuda/tensorrt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:torch`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`tensorrt`](/packages/cuda/tensorrt) [`torch2trt`](/packages/pytorch/torch2trt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:transformers`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`bitsandbytes`](/packages/llm/bitsandbytes) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag vscode)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host vscode:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag vscode)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag vscode) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build vscode\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1526,"sizeBytes":5206},{"name":"README.md","path":"packages/cuda/cuda-python/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cuda/cuda-python/README.md","title":"cuda-python Documentation","category":"plugin-manifest","format":"markdown","content":"# cuda-python\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda-python:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda-python` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cuml`](/packages/ml/rapids/cuml) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss_lite`](/packages/vectordb/faiss_lite) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index:samples`](/packages/rag/llama-index) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`raft`](/packages/ml/rapids/raft) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`tam`](/packages/vit/tam) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.8`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu128`](/packages/cuda/cuda) [`cuda:12.8`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.9`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu129`](/packages/cuda/cuda) [`cuda:12.9`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:13.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu130`](/packages/cuda/cuda) [`cuda:13.0`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/cuda-python:builder-r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-26` | `arm64` | `5.1GB` |\n|   [`dustynv/cuda-python:builder-r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-11` | `arm64` | `3.6GB` |\n|   [`dustynv/cuda-python:r35.2.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.3.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/cuda-python:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cuda-python)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/cuda-python:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/cuda-python:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cuda-python)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cuda-python) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cuda-python\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":5141,"sizeBytes":14359},{"name":"README.md","path":"packages/cuda/cuda/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cuda/cuda/README.md","title":"cuda Documentation","category":"plugin-manifest","format":"markdown","content":"# cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/cuda:12.2-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-05, 3.4GB)`<br>[`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n\n| **`pip_cache:cu122`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2`](/packages/cuda/cuda) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.2-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Images | [`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu124`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4`](/packages/cuda/cuda) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.4-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda` |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`clip_trt`](/packages/vit/clip_trt) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pyav`](/packages/multimedia/pyav) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/packages/speech/whisperx) [`wyoming-piper:1.6.2`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-piper:master`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-whisper:2.5.0`](/packages/smart-home/wyoming/wyoming-whisper) [`wyoming-whisper:master`](/packages/smart-home/wyoming/wyoming-whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xformers:0.0.32`](/packages/attention/xformers) [`xformers:0.0.33`](/packages/attention/xformers) [`xgrammar:0.1.15`](/packages/llm/xgrammar) [`xgrammar:0.1.18`](/packages/llm/xgrammar) [`xgrammar:0.1.19`](/packages/llm/xgrammar) [`xgrammar:0.1.20`](/packages/llm/xgrammar) [`xgrammar:0.1.21`](/packages/llm/xgrammar) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtts`](/packages/speech/xtts) [`xtuner`](/packages/vlm/xtuner) [`zed:5.0`](/packages/hw/zed) [`zed:5.0-humble`](/packages/hw/zed) [`zed:5.0-jazzy`](/packages/hw/zed) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu126`** | |\n| :-- | :-- |\n|    Aliases | `pip_cache` `pip-cache` |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`aim`](/packages/ml/aim) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`canbus`](/packages/hw/canbus) [`cangaroo`](/packages/hw/cangaroo) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`ciso8601:v2.3.2`](/packages/smart-home/dependencies/ciso8601) [`clip_trt`](/packages/vit/clip_trt) [`cmake:pip`](/packages/build/cmake/cmake_pip) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6`](/packages/cuda/cuda) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`docker`](/packages/build/docker) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`graphiti`](/packages/rag/graphiti) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`h5py`](/packages/build/h5py) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`homeassistant-core:2025.7.0`](/packages/smart-home/homeassistant-core) [`huggingface_hub`](/packages/llm/huggingface_hub) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:python`](/packages/multimedia/jetson-utils) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llamaspeak`](/packages/llm/llamaspeak) [`llava`](/packages/vlm/llava) [`llvm:10`](/packages/build/llvm) [`llvm:11`](/packages/build/llvm) [`llvm:12`](/packages/build/llvm) [`llvm:13`](/packages/build/llvm) [`llvm:14`](/packages/build/llvm) [`llvm:15`](/packages/build/llvm) [`llvm:16`](/packages/build/llvm) [`llvm:17`](/packages/build/llvm) [`llvm:18`](/packages/build/llvm) [`llvm:19`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`meshlab:MeshLab-2023.12`](/packages/3d/3dvision/meshlab) [`meshlab:MeshLab-2025.03`](/packages/3d/3dvision/meshlab) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`mujoco`](/packages/sim/mujoco) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`ninja`](/packages/build/ninja) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`numpy`](/packages/numeric/numpy) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnx`](/packages/ml/onnx) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`openai`](/packages/llm/openai) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`polyscope:2.5.0`](/packages/3d/3dvision/polyscope) [`prismatic`](/packages/vlm/prismatic) [`protobuf:apt`](/packages/build/protobuf/protobuf_apt) [`protobuf:cpp`](/packages/build/protobuf/protobuf_cpp) [`protomotions:2.5.0`](/packages/robots/protomotions) [`psutil-home-assistant:0.0.1`](/packages/smart-home/dependencies/psutil-home-assistant) [`pyav`](/packages/multimedia/pyav) [`pybind11`](/packages/build/pybind11) [`pyceres:2.5`](/packages/3d/3dvision/pyceres) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`python:3.10`](/packages/build/python) [`python:3.11`](/packages/build/python) [`python:3.12`](/packages/build/python) [`python:3.13`](/packages/build/python) [`python:3.14`](/packages/build/python) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`riva-client:python`](/packages/speech/riva-client) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`rust`](/packages/build/rust) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim`](/packages/llm/sudonim) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorboard`](/packages/ml/tensorflow/tensorboard) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode`](/packages/code/vscode) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`vtk:9.3.1`](/packages/sim/genesis/vtk) [`vtk:9.4.2`](/packages/sim/genesis/vtk) [`vtk:9.5.0`](/packages/sim/genesis/vtk) [`vulkan:1.3.275`](/packages/multimedia/vulkan) [`vulkan:1.3.296`](/packages/multimedia/vulkan) [`vulkan:1.4.309`](/packages/multimedia/vulkan) [`vulkan:1.4.313`](/packages/multimedia/vulkan) [`vulkan:1.4.314`](/packages/multimedia/vulkan) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/pa\n\n<!-- [TRUNCATED AT 50KB] -->","isInternal":false,"tokens":19059,"sizeBytes":50030},{"name":"README.md","path":"packages/cuda/cudastack/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cuda/cudastack/README.md","title":"cudastack Documentation","category":"plugin-manifest","format":"markdown","content":"# CUDA Stack - Consolidated Library Installation\n\n## What is cudastack?\n\n`cudastack` is a **consolidation package** that installs multiple CUDA libraries (cuDNN, NCCL, TensorRT, CUTLASS, etc.) in **ONE RUN command** to avoid Docker's \"max depth exceeded\" error.\n\n## Important: cudastack vs cuda\n\n```\ncuda/ cudastack/\n├── Dockerfile ├── Dockerfile (ONE RUN for all libs)\n├── Dockerfile.pip ├── config.py (variants)\n├── Dockerfile.builtin └── install/ & build/ scripts\n├── Dockerfile.samples\n└── config.py\n ├── cuda_package() cudastack DEPENDS ON cuda\n ├── cuda_builtin() (doesn't replace it!)\n ├── cuda_samples()\n └── pip_cache()\n```\n\n**Key Points:**\n\n1. ✅ **`cuda` package** - Installs CUDA toolkit, provides pip_cache, samples, builtin variants\n2. ✅ **`cudastack` package** - Builds ON TOP of `cuda`, adds cuDNN/NCCL/TensorRT/etc. in ONE layer\n3. ✅ **pip_cache, cuda_samples, cuda_builtin** - Still work! They're part of `cuda`, not `cudastack`\n\n## The Problem It Solves\n\n### Before (Separate Packages):\n```\nLayer 1: cuda\nLayer 2: cudnn\nLayer 3: tensorrt\nLayer 4: nccl\nLayer 5: cutlass\n...\nLayer 50+: \"max depth exceeded\" ❌\n```\n\n### After (cudastack):\n```\nLayer 1: cuda (with pip_cache, samples, etc.)\nLayer 2: cudastack (cudnn + nccl + tensorrt + cutlass in ONE RUN)\nLayer 3: your-app\nResult: Works! ✅\n```\n\n## Architecture\n\n```\nBASE_IMAGE\n ↓\ncuda:12.4 (from cuda/Dockerfile)\n ├── pip_cache:cu124 (from cuda/Dockerfile.pip)\n └── cuda:12.4-samples (from cuda/Dockerfile.samples)\n ↓\ncudastack:minimal (from cudastack/Dockerfile)\n ├── cuDNN\n ├── NCCL\n └── (all installed in ONE RUN)\n ↓\nyour-app\n```\n\n## Variants\n\n| Variant | Includes | Use Case |\n|---------|----------|----------|\n| `cudastack:minimal` | cuDNN + NCCL | Training, inference |\n| `cudastack:standard` | + TensorRT | Optimized inference |\n| `cudastack:full` | + CUTLASS + GDRCopy | Research, development |\n\n## Usage\n\n### Build with jetson-containers:\n\n```bash\n# Build minimal stack\n./build.sh cudastack:minimal\n\n# Build standard (with TensorRT)\n./build.sh cudastack:standard\n\n# Build full (with CUTLASS)\n./build.sh cudastack:full\n\n# Use as base for your app\n./build.sh --base cudastack:minimal pytorch\n```\n\n### The cudastack Dockerfile does ONE RUN:\n\n```dockerfile\n# Copy all install/build scripts\nCOPY install/*.sh /tmp/cudastack/install/\nCOPY build/*.sh /tmp/cudastack/build/\n\n# ONE RUN command calls all scripts conditionally\nRUN set -ex && \\\n if [ \"$WITH_CUDNN\" = \"1\" ]; then \\\n ./install/install_cudnn.sh; \\\n fi && \\\n if [ \"$WITH_NCCL\" = \"1\" ]; then \\\n ./install/install_nccl.sh; \\\n fi && \\\n if [ \"$WITH_TENSORRT\" = \"1\" ]; then \\\n ./install/install_tensorrt.sh; \\\n fi && \\\n # ... more components ...\n rm -rf /tmp/cudastack\n```\n\n**Key**: All library installations happen in this ONE RUN = ONE Docker layer!\n\n## Setup\n\nThe install scripts need to be populated from existing packages:\n\n```bash\ncd packages/cuda/cudastack\n./setup-scripts.sh\n```\n\nThen review and complete the scripts in `install/` and `build/`.\n\n## What About pip_cache, samples, builtin?\n\nThese are **still available** from the `cuda` package:\n\n```bash\n# pip_cache is automatically included by cuda package\n./build.sh cuda:12.4\n# This creates both:\n# - cuda:12.4 (main CUDA toolkit)\n# - pip_cache:cu124 (pip caching)\n\n# CUDA samples work as before\n./build.sh cuda:12.4-samples\n\n# Builtin works as before (for JetPack 4-5)\n./build.sh cuda:11.4 # Uses Dockerfile.builtin when CUDA pre-installed\n```\n\nThe `cuda` package's `config.py` returns multiple packages:\n```python\n# In cuda/config.py\ndef cuda_package(version, url, ...):\n cuda = package.copy()\n cuda['name'] = f'cuda:{version}'\n \n cuda_pip = pip_cache(version, requires) # Creates pip_cache package\n \n return cuda, cuda_pip # Returns TWO packages!\n```\n\n**cudastack doesn't need to duplicate this** - it just depends on `cuda` being installed first.\n\n## Component Toggles\n\nControlled by `WITH_*` build args (set in config.py):\n\n- ✅ `WITH_CUDNN=1` - cuDNN (always on)\n- ✅ `WITH_NCCL=1` - NCCL (always on)\n- 🎛️ `WITH_TENSORRT` - TensorRT (on in standard/full)\n- 🎛️ `WITH_CUTLASS` - CUTLASS (on in full only)\n- 🎛️ `WITH_GDRCOPY` - GDRCopy (on for Tegra)\n- ⚙️ `WITH_CUDSS=0` - cuDSS (off by default)\n- ⚙️ `WITH_CUSPARSELT=0` - cuSPARSELt (off by default)\n- ⚙️ `WITH_CUTENSOR=0` - cuTENSOR (off by default)\n- ⚙️ `WITH_NVPL=0` - NVPL (off by default)\n- ⚙️ `WITH_NVSHMEM=0` - NVSHMEM (off by default)\n\n## How config.py Works\n\nFollowing the jetson-containers pattern:\n\n```python\ndef cuda_stack(name, with_tensorrt=False, with_cutlass=False, requires=None):\n \"\"\"Generate a cudastack package variant\"\"\"\n pkg = package.copy()\n pkg['name'] = name\n pkg['depends'] = ['cuda'] # Depends ON cuda, doesn't replace it\n \n pkg['build_args'] = {\n 'WITH_CUDNN': '1',\n 'WITH_TENSORRT': '1' if with_tensorrt else '0',\n 'WITH_NCCL': '1',\n # ... etc\n }\n \n return pkg\n\n# Define variants\npackage = [\n cuda_stack('cudastack:minimal', with_tensorrt=False),\n cuda_stack('cudastack:standard', with_tensorrt=True),\n cuda_stack('cudastack:full', with_tensorrt=True, with_cutlass=True),\n]\n```\n\n## Comparison with Individual Packages\n\n| Approach | Layers | Build Time | Complexity |\n|----------|--------|------------|------------|\n| **Individual packages** | 50+ | Longer | Simple per-package |\n| **cudastack (consolidated)** | 2-3 | Faster | One complex layer |\n\nBoth approaches work! Use `cudastack` when you need multiple libraries and want to avoid layer limits.\n\n## See Also\n\n- [cuda package](../cuda/) - CUDA toolkit, pip_cache, samples, builtin\n- [cudnn package](../cudnn/) - Individual cuDNN package\n- [tensorrt package](../tensorrt/) - Individual TensorRT package (has Dockerfile.deb and Dockerfile.tar)\n- [nccl package](../nccl/) - Individual NCCL package\n\n## Summary\n\n✅ **cudastack is a consolidation layer** on top of cuda \n✅ **pip_cache, samples, builtin still work** from the cuda package \n✅ **ONE RUN command** installs all libraries to avoid layer limits \n✅ **Follows jetson-containers patterns** with config.py and variants \n","isInternal":false,"tokens":1829,"sizeBytes":6407},{"name":"README.md","path":"packages/cuda/cutlass/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cuda/cutlass/README.md","title":"cutlass Documentation","category":"plugin-manifest","format":"markdown","content":"# cutlass\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cutlass:4.1.0`** | |\n| :-- | :-- |\n|    Aliases | `cutlass` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cutlass)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cutlass:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cutlass)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cutlass) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cutlass\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":2912,"sizeBytes":8228},{"name":"README.md","path":"packages/cuda/pycuda/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cuda/pycuda/README.md","title":"pycuda Documentation","category":"plugin-manifest","format":"markdown","content":"# pycuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 0.4GB)`<br>[`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-09-07, 5.0GB)`<br>[`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-08-29, 5.0GB)`<br>[`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 5.0GB)`<br>[`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 3.5GB)`<br>[`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2025-03-03, 2.3GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `0.4GB` |\n|   [`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-09-07` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycuda)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/pycuda:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/pycuda:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":2214,"sizeBytes":6522},{"name":"README.md","path":"packages/cv/3d/3dobjects/partpacker/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dobjects/partpacker/README.md","title":"partpacker Documentation","category":"plugin-manifest","format":"markdown","content":"# partpacker\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`partpacker:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `partpacker` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/PartPacker |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag partpacker)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host partpacker:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag partpacker)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag partpacker) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build partpacker\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1124,"sizeBytes":3918},{"name":"README.md","path":"packages/cv/3d/3dobjects/sparc3d/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dobjects/sparc3d/README.md","title":"sparc3d Documentation","category":"plugin-manifest","format":"markdown","content":"# sparc3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparc3d:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparc3d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/lizhihao6/Sparc3D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparc3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparc3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparc3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparc3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparc3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1136,"sizeBytes":3894},{"name":"README.md","path":"packages/cv/3d/3dvision/glomap/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/glomap/README.md","title":"glomap Documentation","category":"plugin-manifest","format":"markdown","content":"# glomap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`glomap:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `glomap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag glomap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host glomap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag glomap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag glomap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build glomap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":993,"sizeBytes":3413},{"name":"README.md","path":"packages/cv/3d/3dvision/hloc/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/hloc/README.md","title":"hloc Documentation","category":"plugin-manifest","format":"markdown","content":"# hloc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`hloc:1.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n| **`hloc:1.5`** | |\n| :-- | :-- |\n|    Aliases | `hloc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag hloc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host hloc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag hloc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag hloc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build hloc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1404,"sizeBytes":4662},{"name":"README.md","path":"packages/cv/3d/3dvision/kaolin/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/kaolin/README.md","title":"kaolin Documentation","category":"plugin-manifest","format":"markdown","content":"# kaolin\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`kaolin:0.18.0`** | |\n| :-- | :-- |\n|    Aliases | `kaolin` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVIDIAGameWorks/kaolin |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag kaolin)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host kaolin:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag kaolin)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag kaolin) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build kaolin\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":965,"sizeBytes":3329},{"name":"README.md","path":"packages/cv/3d/3dvision/meshlab/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/meshlab/README.md","title":"meshlab Documentation","category":"plugin-manifest","format":"markdown","content":"# meshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`meshlab:MeshLab-2023.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n| **`meshlab:MeshLab-2025.03`** | |\n| :-- | :-- |\n|    Aliases | `meshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag meshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host meshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag meshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag meshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build meshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":981,"sizeBytes":3439},{"name":"README.md","path":"packages/cv/3d/3dvision/nerfacc/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/nerfacc/README.md","title":"nerfacc Documentation","category":"plugin-manifest","format":"markdown","content":"# nerfacc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfacc:0.5.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n| **`nerfacc:0.5.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfacc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfacc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfacc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfacc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfacc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfacc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1178,"sizeBytes":3942},{"name":"README.md","path":"packages/cv/3d/3dvision/open3d/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/open3d/README.md","title":"open3d Documentation","category":"plugin-manifest","format":"markdown","content":"# open3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`open3d:1.19.0`** | |\n| :-- | :-- |\n|    Aliases | `open3d` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag open3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host open3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag open3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag open3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build open3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1039,"sizeBytes":3557},{"name":"README.md","path":"packages/cv/3d/3dvision/pixsfm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/pixsfm/README.md","title":"pixsfm Documentation","category":"plugin-manifest","format":"markdown","content":"# pixsfm\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pixsfm:1.0`** | |\n| :-- | :-- |\n|    Aliases | `pixsfm` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pixsfm)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pixsfm:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pixsfm)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pixsfm) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pixsfm\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1004,"sizeBytes":3492},{"name":"README.md","path":"packages/cv/3d/3dvision/polyscope/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/polyscope/README.md","title":"polyscope Documentation","category":"plugin-manifest","format":"markdown","content":"# polyscope\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`polyscope:2.5.0`** | |\n| :-- | :-- |\n|    Aliases | `polyscope` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nmwsharp/polyscope-py.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag polyscope)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host polyscope:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag polyscope)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag polyscope) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build polyscope\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1101,"sizeBytes":3603},{"name":"README.md","path":"packages/cv/3d/3dvision/pyceres/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/pyceres/README.md","title":"pyceres Documentation","category":"plugin-manifest","format":"markdown","content":"# pyceres\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pyceres:2.5`** | |\n| :-- | :-- |\n|    Aliases | `pyceres` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`ninja`](/packages/build/ninja) [`cmake`](/packages/build/cmake/cmake_pip) [`opengl`](/packages/multimedia/opengl) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/pyceres.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pyceres)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pyceres:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pyceres)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pyceres) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pyceres\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1057,"sizeBytes":3469},{"name":"README.md","path":"packages/cv/3d/3dvision/pycolmap/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/pycolmap/README.md","title":"pycolmap Documentation","category":"plugin-manifest","format":"markdown","content":"# pycolmap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycolmap:3.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n| **`pycolmap:3.13`** | |\n| :-- | :-- |\n|    Aliases | `pycolmap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycolmap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pycolmap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycolmap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycolmap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycolmap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1395,"sizeBytes":4503},{"name":"README.md","path":"packages/cv/3d/3dvision/pymeshlab/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/pymeshlab/README.md","title":"pymeshlab Documentation","category":"plugin-manifest","format":"markdown","content":"# pymeshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pymeshlab:2023.12.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2023.12.post3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n| **`pymeshlab:2025.6.23.dev0`** | |\n| :-- | :-- |\n|    Aliases | `pymeshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`genesis-world:0.2.2`](/packages/sim/genesis) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`protomotions:2.5.0`](/packages/robots/protomotions) [`robogen`](/packages/sim/robogen) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/PyMeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pymeshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host pymeshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pymeshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pymeshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pymeshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1417,"sizeBytes":4608},{"name":"README.md","path":"packages/cv/3d/3dvision/tinycudann/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/3dvision/tinycudann/README.md","title":"tinycudann Documentation","category":"plugin-manifest","format":"markdown","content":"# tinycudann\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`tinycudann:1.7`** | |\n| :-- | :-- |\n|    Aliases | `tinycudann` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/tiny-cuda-nn |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag tinycudann)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host tinycudann:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag tinycudann)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag tinycudann) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build tinycudann\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1111,"sizeBytes":3621},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/3dgrut/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/3dgrut/README.md","title":"3dgrut Documentation","category":"plugin-manifest","format":"markdown","content":"# 3dgrut\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3dgrut:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `3dgrut` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cupy`](/packages/numeric/cupy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`kaolin`](/packages/3d/3dvision/kaolin) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nv-tlabs/3dgrut |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3dgrut)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 3dgrut:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3dgrut)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3dgrut) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3dgrut\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1060,"sizeBytes":3589},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/4k4d/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/4k4d/README.md","title":"4k4d Documentation","category":"plugin-manifest","format":"markdown","content":"# 4k4d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`4k4d:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `4k4d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) [`easyvolcap`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/4k4D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 4k4d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host 4k4d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 4k4d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 4k4d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 4k4d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1081,"sizeBytes":3634},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/easyvolcap/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/easyvolcap/README.md","title":"easyvolcap Documentation","category":"plugin-manifest","format":"markdown","content":"# easyvolcap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`easyvolcap:0.0.0`** | |\n| :-- | :-- |\n|    Aliases | `easyvolcap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) [`fast_gauss`](/packages/3d/gaussian_splatting/fast_gauss) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/zju3dv/EasyVolcap |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag easyvolcap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host easyvolcap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag easyvolcap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag easyvolcap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build easyvolcap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1086,"sizeBytes":3717},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/fast_gauss/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/fast_gauss/README.md","title":"fast_gauss Documentation","category":"plugin-manifest","format":"markdown","content":"# fast_gauss\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fast_gauss:1.0.0`** | |\n| :-- | :-- |\n|    Aliases | `fast_gauss` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`open3d`](/packages/3d/3dvision/open3d) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dependants | [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/dendenxu/fast-gaussian-rasterization |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fast_gauss)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host fast_gauss:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fast_gauss)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fast_gauss) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fast_gauss\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1095,"sizeBytes":3742},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/gsplat/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/gsplat/README.md","title":"gsplat Documentation","category":"plugin-manifest","format":"markdown","content":"# gsplat\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`gsplat:1.5.3`** | |\n| :-- | :-- |\n|    Aliases | `gsplat` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/gsplat |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag gsplat)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host gsplat:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag gsplat)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag gsplat) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build gsplat\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1049,"sizeBytes":3570},{"name":"README.md","path":"packages/cv/3d/gaussian_splatting/nerfview/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/gaussian_splatting/nerfview/README.md","title":"nerfview Documentation","category":"plugin-manifest","format":"markdown","content":"# nerfview\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfview:0.1.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfview` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`torch`](/packages/pytorch) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`tinycudann`](/packages/3d/3dvision/tinycudann) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://pypi.org/project/nerfview/ |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfview)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfview:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfview)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfview) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfview\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1060,"sizeBytes":3578},{"name":"README.md","path":"packages/cv/3d/nerf/fruitnerf/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/nerf/fruitnerf/README.md","title":"fruitnerf Documentation","category":"plugin-manifest","format":"markdown","content":"# fruitnerf\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`fruitnerf:1.0`** | |\n| :-- | :-- |\n|    Aliases | `fruitnerf` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) [`nerfstudio`](/packages/3d/nerf/nerfstudio) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/meyerls/FruitNeRF |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/fruitnerf:r36.3.0`](https://hub.docker.com/r/dustynv/fruitnerf/tags) | `2024-09-02` | `arm64` | `14.6GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag fruitnerf)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/fruitnerf:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/fruitnerf:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag fruitnerf)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag fruitnerf) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build fruitnerf\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1549,"sizeBytes":5029},{"name":"README.md","path":"packages/cv/3d/nerf/nerfstudio/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/3d/nerf/nerfstudio/README.md","title":"nerfstudio Documentation","category":"plugin-manifest","format":"markdown","content":"# nerfstudio\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfstudio:1.1.7`** | |\n| :-- | :-- |\n|    Aliases | `nerfstudio` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`h5py`](/packages/build/h5py) [`ninja`](/packages/build/ninja) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`torchaudio`](/packages/pytorch/torchaudio) [`nerfacc`](/packages/3d/3dvision/nerfacc) [`torch`](/packages/pytorch) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`glomap`](/packages/3d/3dvision/glomap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) [`hloc`](/packages/3d/3dvision/hloc) [`nerfview`](/packages/3d/gaussian_splatting/nerfview) [`gsplat`](/packages/3d/gaussian_splatting/gsplat) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfstudio |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/nerfstudio:r36.3.0`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2024-09-02` | `arm64` | `7.7GB` |\n|   [`dustynv/nerfstudio:r36.4.0-cu128`](https://hub.docker.com/r/dustynv/nerfstudio/tags) | `2025-02-11` | `arm64` | `7.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfstudio)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/nerfstudio:r36.4.0-cu128\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/nerfstudio:r36.4.0-cu128\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfstudio)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfstudio) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfstudio\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1610,"sizeBytes":5156},{"name":"README.md","path":"packages/cv/cv-cuda/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/cv-cuda/README.md","title":"cv-cuda Documentation","category":"plugin-manifest","format":"markdown","content":"# cv-cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cv-cuda:0.15`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15` `cv-cuda` `cvcuda` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg:git`](/packages/multimedia/ffmpeg) [`pyav`](/packages/multimedia/pyav) [`pycuda`](/packages/cuda/pycuda) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cv-cuda:0.15-cpp`** | |\n| :-- | :-- |\n|    Aliases | `cvcuda:0.15-cpp` `cv-cuda:cpp` `cvcuda:cpp` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) |\n|    Dependants | [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cv-cuda)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cv-cuda:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cv-cuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cv-cuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cv-cuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1498,"sizeBytes":4766},{"name":"README.md","path":"packages/cv/deepstream/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/deepstream/README.md","title":"deepstream Documentation","category":"plugin-manifest","format":"markdown","content":"# deepstream\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`deepstream`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`cmake`](/packages/build/cmake/cmake_pip) [`numpy`](/packages/numeric/numpy) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`gstreamer`](/packages/multimedia/gstreamer) [`tritonserver`](/packages/ml/tritonserver) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-06, 2.3GB)`<br>[`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-22, 6.8GB)`<br>[`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 6.8GB)`<br>[`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2023-12-06, 6.7GB)`<br>[`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) `(2024-03-05, 9.8GB)` |\n|    Notes | https://docs.nvidia.com/metropolis/deepstream/dev-guide/text/DS_Quickstart.html |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/deepstream:r32.7.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-06` | `arm64` | `2.3GB` |\n|   [`dustynv/deepstream:r35.2.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-22` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.3.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `6.8GB` |\n|   [`dustynv/deepstream:r35.4.1`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2023-12-06` | `arm64` | `6.7GB` |\n|   [`dustynv/deepstream:r36.2.0`](https://hub.docker.com/r/dustynv/deepstream/tags) | `2024-03-05` | `arm64` | `9.8GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag deepstream)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/deepstream:r32.7.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/deepstream:r32.7.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag deepstream)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag deepstream) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build deepstream\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1658,"sizeBytes":5165},{"name":"README.md","path":"packages/cv/diffusion/3d_diffusion_policy/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/3d_diffusion_policy/README.md","title":"3d_diffusion_policy Documentation","category":"plugin-manifest","format":"markdown","content":"# 3d_diffusion_policy\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n\n3D Diffusion Policy: Generalizable Visuomotor Policy Learning (https://github.com/YanjieZe/3D-Diffusion-Policy)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`3d_diffusion_policy`** | |\n| :-- | :-- |\n|    Aliases | `3d-diffusion-policy` |\n|    Requires | `L4T ['>=36']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`mujoco`](/packages/sim/mujoco) [`numba`](/packages/numeric/numba) [`diffusers`](/packages/diffusion/diffusers) [`torchaudio`](/packages/pytorch/torchaudio) [`pytorch3d`](/packages/pytorch/torch3d) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) `(2024-09-30, 7.0GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/3d_diffusion_policy:r36.3.0`](https://hub.docker.com/r/dustynv/3d_diffusion_policy/tags) | `2024-09-30` | `arm64` | `7.0GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag 3d_diffusion_policy)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/3d_diffusion_policy:r36.3.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/3d_diffusion_policy:r36.3.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag 3d_diffusion_policy)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag 3d_diffusion_policy) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build 3d_diffusion_policy\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":1325,"sizeBytes":4440},{"name":"README.md","path":"packages/cv/diffusion/ai-toolkit/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/ai-toolkit/README.md","title":"ai-toolkit Documentation","category":"plugin-manifest","format":"markdown","content":"# ai-toolkit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\ndocs.md\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`ai-toolkit`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`bitsandbytes`](/packages/llm/bitsandbytes) [`diffusers`](/packages/diffusion/diffusers) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ai-toolkit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ai-toolkit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ai-toolkit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ai-toolkit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ai-toolkit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":943,"sizeBytes":3350},{"name":"README.md","path":"packages/cv/diffusion/cache_edit/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/cache_edit/README.md","title":"cache_edit Documentation","category":"plugin-manifest","format":"markdown","content":"# cache_edit\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cache_dit:0.2.0`** | |\n| :-- | :-- |\n|    Aliases | `cache_dit` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/vipshop/cache-dit |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cache_edit)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cache_edit:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cache_edit)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cache_edit) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cache_edit\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n","isInternal":false,"tokens":920,"sizeBytes":3284},{"name":"README.md","path":"packages/cv/diffusion/comfyui/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/comfyui/README.md","title":"comfyui Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fcomfyui%2FREADME.md -->","isInternal":false,"tokens":1863,"sizeBytes":5846},{"name":"README.md","path":"packages/cv/diffusion/diffusers/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/diffusers/README.md","title":"diffusers Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusers%2FREADME.md -->","isInternal":false,"tokens":3285,"sizeBytes":9354},{"name":"README.md","path":"packages/cv/diffusion/diffusion_policy/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/diffusion_policy/README.md","title":"diffusion_policy Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fdiffusion_policy%2FREADME.md -->","isInternal":false,"tokens":1430,"sizeBytes":4856},{"name":"README.md","path":"packages/cv/diffusion/fastgen/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/fastgen/README.md","title":"fastgen Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Ffastgen%2FREADME.md -->","isInternal":false,"tokens":879,"sizeBytes":3304},{"name":"README.md","path":"packages/cv/diffusion/framepack/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/framepack/README.md","title":"framepack Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fframepack%2FREADME.md -->","isInternal":false,"tokens":1110,"sizeBytes":3921},{"name":"README.md","path":"packages/cv/diffusion/sdnext/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/sdnext/README.md","title":"sdnext Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fsdnext%2FREADME.md -->","isInternal":false,"tokens":1150,"sizeBytes":4058},{"name":"README.md","path":"packages/cv/diffusion/self-forcing/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/self-forcing/README.md","title":"self-forcing Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fself-forcing%2FREADME.md -->","isInternal":false,"tokens":1104,"sizeBytes":3914},{"name":"README.md","path":"packages/cv/diffusion/stable-diffusion-webui/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/diffusion/stable-diffusion-webui/README.md","title":"stable-diffusion-webui Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fdiffusion%2Fstable-diffusion-webui%2FREADME.md -->","isInternal":false,"tokens":2449,"sizeBytes":7905},{"name":"README.md","path":"packages/cv/holoscan/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/holoscan/README.md","title":"holoscan Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fholoscan%2FREADME.md -->","isInternal":false,"tokens":1189,"sizeBytes":3970},{"name":"README.md","path":"packages/cv/jetson-inference/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/jetson-inference/README.md","title":"jetson-inference Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fjetson-inference%2FREADME.md -->","isInternal":false,"tokens":5116,"sizeBytes":14579},{"name":"README.md","path":"packages/cv/opencv/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/opencv/README.md","title":"opencv Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fopencv%2FREADME.md -->","isInternal":false,"tokens":6627,"sizeBytes":18954},{"name":"README.md","path":"packages/cv/vpi/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/cv/vpi/README.md","title":"vpi Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fcv%2Fvpi%2FREADME.md -->","isInternal":false,"tokens":1340,"sizeBytes":4307},{"name":"README.md","path":"packages/hw/canable/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/canable/README.md","title":"canable Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanable%2FREADME.md -->","isInternal":false,"tokens":749,"sizeBytes":2753},{"name":"README.md","path":"packages/hw/canbus/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/canbus/README.md","title":"canbus Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcanbus%2FREADME.md -->","isInternal":false,"tokens":767,"sizeBytes":2813},{"name":"README.md","path":"packages/hw/cangaroo/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/cangaroo/README.md","title":"cangaroo Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fcangaroo%2FREADME.md -->","isInternal":false,"tokens":782,"sizeBytes":2844},{"name":"README.md","path":"packages/hw/jetcam/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/jetcam/README.md","title":"jetcam Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjetcam%2FREADME.md -->","isInternal":false,"tokens":918,"sizeBytes":3262},{"name":"README.md","path":"packages/hw/jupyter_clickable_image_widget/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/jupyter_clickable_image_widget/README.md","title":"jupyter_clickable_image_widget Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fjupyter_clickable_image_widget%2FREADME.md -->","isInternal":false,"tokens":909,"sizeBytes":3311},{"name":"README.md","path":"packages/hw/librealsense/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/librealsense/README.md","title":"librealsense Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Flibrealsense%2FREADME.md -->","isInternal":false,"tokens":1525,"sizeBytes":4698},{"name":"README.md","path":"packages/hw/oled/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/oled/README.md","title":"oled Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Foled%2FREADME.md -->","isInternal":false,"tokens":718,"sizeBytes":2609},{"name":"README.md","path":"packages/hw/pl2303/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/pl2303/README.md","title":"pl2303 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fpl2303%2FREADME.md -->","isInternal":false,"tokens":735,"sizeBytes":2690},{"name":"README.md","path":"packages/hw/zed/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/hw/zed/README.md","title":"zed Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fhw%2Fzed%2FREADME.md -->","isInternal":false,"tokens":2016,"sizeBytes":6362},{"name":"README.md","path":"packages/llm/awq/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/awq/README.md","title":"awq Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fawq%2FREADME.md -->","isInternal":false,"tokens":2927,"sizeBytes":8465},{"name":"README.md","path":"packages/llm/bitsandbytes/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/bitsandbytes/README.md","title":"bitsandbytes Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fbitsandbytes%2FREADME.md -->","isInternal":false,"tokens":3507,"sizeBytes":10552},{"name":"README.md","path":"packages/llm/deepspeed/deepspeed-kernels/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/deepspeed/deepspeed-kernels/README.md","title":"deepspeed-kernels Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2Fdeepspeed-kernels%2FREADME.md -->","isInternal":false,"tokens":1652,"sizeBytes":4838},{"name":"README.md","path":"packages/llm/deepspeed/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/deepspeed/README.md","title":"deepspeed Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdeepspeed%2FREADME.md -->","isInternal":false,"tokens":2249,"sizeBytes":6676},{"name":"README.md","path":"packages/llm/dynamo/dynamo/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/dynamo/dynamo/README.md","title":"dynamo Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fdynamo%2FREADME.md -->","isInternal":false,"tokens":1249,"sizeBytes":4259},{"name":"README.md","path":"packages/llm/dynamo/kai-scheduler/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/dynamo/kai-scheduler/README.md","title":"kai-scheduler Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fkai-scheduler%2FREADME.md -->","isInternal":false,"tokens":850,"sizeBytes":3066},{"name":"README.md","path":"packages/llm/dynamo/mooncake/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/dynamo/mooncake/README.md","title":"mooncake Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fmooncake%2FREADME.md -->","isInternal":false,"tokens":1094,"sizeBytes":3728},{"name":"README.md","path":"packages/llm/dynamo/nixl/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/dynamo/nixl/README.md","title":"nixl Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fdynamo%2Fnixl%2FREADME.md -->","isInternal":false,"tokens":1108,"sizeBytes":3729},{"name":"README.md","path":"packages/llm/exllama/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/exllama/README.md","title":"exllama Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fexllama%2FREADME.md -->","isInternal":false,"tokens":2218,"sizeBytes":6973},{"name":"README.md","path":"packages/llm/gptqmodel/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/gptqmodel/README.md","title":"gptqmodel Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fgptqmodel%2FREADME.md -->","isInternal":false,"tokens":1090,"sizeBytes":3806},{"name":"README.md","path":"packages/llm/huggingface_hub/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/huggingface_hub/README.md","title":"huggingface_hub Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhuggingface_hub%2FREADME.md -->","isInternal":false,"tokens":5320,"sizeBytes":14449},{"name":"README.md","path":"packages/llm/hymba/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/hymba/README.md","title":"hymba Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fhymba%2FREADME.md -->","isInternal":false,"tokens":1009,"sizeBytes":3539},{"name":"README.md","path":"packages/llm/ktransformers/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/ktransformers/README.md","title":"ktransformers Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fktransformers%2FREADME.md -->","isInternal":false,"tokens":918,"sizeBytes":3296},{"name":"README.md","path":"packages/llm/llama_cpp/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/llama_cpp/README.md","title":"llama_cpp Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama_cpp%2FREADME.md -->","isInternal":false,"tokens":6663,"sizeBytes":18458},{"name":"README.md","path":"packages/llm/llama-factory/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/llama-factory/README.md","title":"llama-factory Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllama-factory%2FREADME.md -->","isInternal":false,"tokens":1672,"sizeBytes":5398},{"name":"README.md","path":"packages/llm/llamaspeak/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/llamaspeak/README.md","title":"llamaspeak Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fllamaspeak%2FREADME.md -->","isInternal":false,"tokens":2447,"sizeBytes":8441},{"name":"README.md","path":"packages/llm/lobe_chat/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/lobe_chat/README.md","title":"lobe_chat Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flobe_chat%2FREADME.md -->","isInternal":false,"tokens":1114,"sizeBytes":3969},{"name":"README.md","path":"packages/llm/local_llm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/local_llm/README.md","title":"local_llm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Flocal_llm%2FREADME.md -->","isInternal":false,"tokens":6461,"sizeBytes":21386},{"name":"README.md","path":"packages/llm/minference/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/minference/README.md","title":"minference Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fminference%2FREADME.md -->","isInternal":false,"tokens":1324,"sizeBytes":4325},{"name":"README.md","path":"packages/llm/mistral/mistral_rs/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/mistral/mistral_rs/README.md","title":"mistral_rs Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmistral%2Fmistral_rs%2FREADME.md -->","isInternal":false,"tokens":771,"sizeBytes":2814},{"name":"README.md","path":"packages/llm/mlc/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/mlc/README.md","title":"mlc Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fmlc%2FREADME.md -->","isInternal":false,"tokens":7009,"sizeBytes":18568},{"name":"README.md","path":"packages/llm/nano_llm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/nano_llm/README.md","title":"nano_llm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnano_llm%2FREADME.md -->","isInternal":false,"tokens":16963,"sizeBytes":50030},{"name":"README.md","path":"packages/llm/nemo/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/nemo/README.md","title":"nemo Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fnemo%2FREADME.md -->","isInternal":false,"tokens":1637,"sizeBytes":4962},{"name":"README.md","path":"packages/llm/ollama/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/ollama/README.md","title":"ollama Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Follama%2FREADME.md -->","isInternal":false,"tokens":4074,"sizeBytes":12148},{"name":"README.md","path":"packages/llm/open-webui/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/open-webui/README.md","title":"open-webui Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopen-webui%2FREADME.md -->","isInternal":false,"tokens":250,"sizeBytes":842},{"name":"README.md","path":"packages/llm/openai/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/openai/README.md","title":"openai Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fopenai%2FREADME.md -->","isInternal":false,"tokens":775,"sizeBytes":2840},{"name":"README.md","path":"packages/llm/optimum/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/optimum/README.md","title":"optimum Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Foptimum%2FREADME.md -->","isInternal":false,"tokens":1508,"sizeBytes":4762},{"name":"README.md","path":"packages/llm/sglang/genai-bench/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/sglang/genai-bench/README.md","title":"genai-bench Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fgenai-bench%2FREADME.md -->","isInternal":false,"tokens":1195,"sizeBytes":4021},{"name":"README.md","path":"packages/llm/sglang/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/sglang/README.md","title":"sglang Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2FREADME.md -->","isInternal":false,"tokens":2872,"sizeBytes":8936},{"name":"README.md","path":"packages/llm/sglang/sgl-kernel/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/sglang/sgl-kernel/README.md","title":"sgl-kernel Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsglang%2Fsgl-kernel%2FREADME.md -->","isInternal":false,"tokens":1192,"sizeBytes":4015},{"name":"README.md","path":"packages/llm/sudonim/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/sudonim/README.md","title":"sudonim Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fsudonim%2FREADME.md -->","isInternal":false,"tokens":2461,"sizeBytes":6994},{"name":"README.md","path":"packages/llm/tensorrt_optimizer/nvidia-modelopt/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/tensorrt_optimizer/nvidia-modelopt/README.md","title":"nvidia-modelopt Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Fnvidia-modelopt%2FREADME.md -->","isInternal":false,"tokens":1054,"sizeBytes":3679},{"name":"README.md","path":"packages/llm/tensorrt_optimizer/tensorrt_llm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/tensorrt_optimizer/tensorrt_llm/README.md","title":"tensorrt_llm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftensorrt_optimizer%2Ftensorrt_llm%2FREADME.md -->","isInternal":false,"tokens":1802,"sizeBytes":5797},{"name":"README.md","path":"packages/llm/text-generation-inference/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/text-generation-inference/README.md","title":"text-generation-inference Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-inference%2FREADME.md -->","isInternal":false,"tokens":1453,"sizeBytes":4909},{"name":"README.md","path":"packages/llm/text-generation-webui/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/text-generation-webui/README.md","title":"text-generation-webui Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftext-generation-webui%2FREADME.md -->","isInternal":false,"tokens":3729,"sizeBytes":12605},{"name":"README.md","path":"packages/llm/transformers/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/transformers/README.md","title":"transformers Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Ftransformers%2FREADME.md -->","isInternal":false,"tokens":5954,"sizeBytes":16981},{"name":"README.md","path":"packages/llm/vllm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/vllm/README.md","title":"vllm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fvllm%2FREADME.md -->","isInternal":false,"tokens":4787,"sizeBytes":14256},{"name":"README.md","path":"packages/llm/xgrammar/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/llm/xgrammar/README.md","title":"xgrammar Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fllm%2Fxgrammar%2FREADME.md -->","isInternal":false,"tokens":2372,"sizeBytes":7360},{"name":"README.md","path":"packages/ml/aim/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/aim/README.md","title":"aim Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Faim%2FREADME.md -->","isInternal":false,"tokens":1072,"sizeBytes":3581},{"name":"README.md","path":"packages/ml/apache/tvm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/apache/tvm/README.md","title":"tvm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fapache%2Ftvm%2FREADME.md -->","isInternal":false,"tokens":808,"sizeBytes":2895},{"name":"README.md","path":"packages/ml/ctranslate2/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/ctranslate2/README.md","title":"ctranslate2 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fctranslate2%2FREADME.md -->","isInternal":false,"tokens":1075,"sizeBytes":3711},{"name":"README.md","path":"packages/ml/dli/dli-nano-ai/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/dli/dli-nano-ai/README.md","title":"dli-nano-ai Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fdli%2Fdli-nano-ai%2FREADME.md -->","isInternal":false,"tokens":1065,"sizeBytes":3708},{"name":"README.md","path":"packages/ml/jax/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/jax/README.md","title":"jax Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fjax%2FREADME.md -->","isInternal":false,"tokens":1597,"sizeBytes":4920},{"name":"README.md","path":"packages/ml/kans/kat/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/kans/kat/README.md","title":"kat Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fkat%2FREADME.md -->","isInternal":false,"tokens":892,"sizeBytes":3175},{"name":"README.md","path":"packages/ml/kans/pykan/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/kans/pykan/README.md","title":"pykan Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fkans%2Fpykan%2FREADME.md -->","isInternal":false,"tokens":880,"sizeBytes":3120},{"name":"README.md","path":"packages/ml/l4t/l4t-diffusion/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-diffusion/README.md","title":"l4t-diffusion Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-diffusion%2FREADME.md -->","isInternal":false,"tokens":1949,"sizeBytes":6301},{"name":"README.md","path":"packages/ml/l4t/l4t-dynamo/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-dynamo/README.md","title":"l4t-dynamo Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-dynamo%2FREADME.md -->","isInternal":false,"tokens":1215,"sizeBytes":4144},{"name":"README.md","path":"packages/ml/l4t/l4t-ml/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-ml/README.md","title":"l4t-ml Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-ml%2FREADME.md -->","isInternal":false,"tokens":2010,"sizeBytes":5839},{"name":"README.md","path":"packages/ml/l4t/l4t-pytorch/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-pytorch/README.md","title":"l4t-pytorch Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-pytorch%2FREADME.md -->","isInternal":false,"tokens":2118,"sizeBytes":5917},{"name":"README.md","path":"packages/ml/l4t/l4t-tensorflow/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-tensorflow/README.md","title":"l4t-tensorflow Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-tensorflow%2FREADME.md -->","isInternal":false,"tokens":2574,"sizeBytes":7300},{"name":"README.md","path":"packages/ml/l4t/l4t-text-generation/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/l4t/l4t-text-generation/README.md","title":"l4t-text-generation Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fl4t%2Fl4t-text-generation%2FREADME.md -->","isInternal":false,"tokens":1737,"sizeBytes":5506},{"name":"README.md","path":"packages/ml/mamba/causalconv1d/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/causalconv1d/README.md","title":"causalconv1d Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcausalconv1d%2FREADME.md -->","isInternal":false,"tokens":1347,"sizeBytes":4209},{"name":"README.md","path":"packages/ml/mamba/cobra/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/cobra/README.md","title":"cobra Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fcobra%2FREADME.md -->","isInternal":false,"tokens":959,"sizeBytes":3364},{"name":"README.md","path":"packages/ml/mamba/dimba/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/dimba/README.md","title":"dimba Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fdimba%2FREADME.md -->","isInternal":false,"tokens":973,"sizeBytes":3402},{"name":"README.md","path":"packages/ml/mamba/mamba/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/mamba/README.md","title":"mamba Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmamba%2FREADME.md -->","isInternal":false,"tokens":1696,"sizeBytes":5170},{"name":"README.md","path":"packages/ml/mamba/mambavision/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/mambavision/README.md","title":"mambavision Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fmambavision%2FREADME.md -->","isInternal":false,"tokens":1238,"sizeBytes":4168},{"name":"README.md","path":"packages/ml/mamba/videomambasuite/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/videomambasuite/README.md","title":"videomambasuite Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fvideomambasuite%2FREADME.md -->","isInternal":false,"tokens":982,"sizeBytes":3453},{"name":"README.md","path":"packages/ml/mamba/zigma/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/mamba/zigma/README.md","title":"zigma Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fmamba%2Fzigma%2FREADME.md -->","isInternal":false,"tokens":1052,"sizeBytes":3688},{"name":"README.md","path":"packages/ml/numeric/arrow/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/numeric/arrow/README.md","title":"arrow Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Farrow%2FREADME.md -->","isInternal":false,"tokens":1464,"sizeBytes":4843},{"name":"README.md","path":"packages/ml/numeric/cupy/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/numeric/cupy/README.md","title":"cupy Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fcupy%2FREADME.md -->","isInternal":false,"tokens":1877,"sizeBytes":5358},{"name":"README.md","path":"packages/ml/numeric/numba/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/numeric/numba/README.md","title":"numba Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumba%2FREADME.md -->","isInternal":false,"tokens":2665,"sizeBytes":7260},{"name":"README.md","path":"packages/ml/numeric/numpy/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/numeric/numpy/README.md","title":"numpy Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fnumpy%2FREADME.md -->","isInternal":false,"tokens":9612,"sizeBytes":25689},{"name":"README.md","path":"packages/ml/numeric/warp/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/numeric/warp/README.md","title":"warp Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fnumeric%2Fwarp%2FREADME.md -->","isInternal":false,"tokens":2867,"sizeBytes":8913},{"name":"README.md","path":"packages/ml/onnx/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/onnx/README.md","title":"onnx Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnx%2FREADME.md -->","isInternal":false,"tokens":7231,"sizeBytes":19344},{"name":"README.md","path":"packages/ml/onnxruntime_genai/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/onnxruntime_genai/README.md","title":"onnxruntime_genai Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime_genai%2FREADME.md -->","isInternal":false,"tokens":949,"sizeBytes":3363},{"name":"README.md","path":"packages/ml/onnxruntime/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/onnxruntime/README.md","title":"onnxruntime Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fonnxruntime%2FREADME.md -->","isInternal":false,"tokens":3805,"sizeBytes":10716},{"name":"README.md","path":"packages/ml/pytorch/apex/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/apex/README.md","title":"apex Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Fapex%2FREADME.md -->","isInternal":false,"tokens":974,"sizeBytes":3435},{"name":"README.md","path":"packages/ml/pytorch/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/README.md","title":"pytorch Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2FREADME.md -->","isInternal":false,"tokens":14601,"sizeBytes":39985},{"name":"README.md","path":"packages/ml/pytorch/torch_tensorrt/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torch_tensorrt/README.md","title":"torch_tensorrt Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch_tensorrt%2FREADME.md -->","isInternal":false,"tokens":1522,"sizeBytes":4831},{"name":"README.md","path":"packages/ml/pytorch/torch2trt/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torch2trt/README.md","title":"torch2trt Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch2trt%2FREADME.md -->","isInternal":false,"tokens":2777,"sizeBytes":7496},{"name":"README.md","path":"packages/ml/pytorch/torch3d/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torch3d/README.md","title":"torch3d Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorch3d%2FREADME.md -->","isInternal":false,"tokens":1419,"sizeBytes":4605},{"name":"README.md","path":"packages/ml/pytorch/torchao/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchao/README.md","title":"torchao Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchao%2FREADME.md -->","isInternal":false,"tokens":2357,"sizeBytes":6904},{"name":"README.md","path":"packages/ml/pytorch/torchaudio/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchaudio/README.md","title":"torchaudio Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchaudio%2FREADME.md -->","isInternal":false,"tokens":5860,"sizeBytes":15819},{"name":"README.md","path":"packages/ml/pytorch/torchsaver/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchsaver/README.md","title":"torchsaver Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsaver%2FREADME.md -->","isInternal":false,"tokens":1166,"sizeBytes":3760},{"name":"README.md","path":"packages/ml/pytorch/torchsde/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchsde/README.md","title":"torchsde Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchsde%2FREADME.md -->","isInternal":false,"tokens":2182,"sizeBytes":6817},{"name":"README.md","path":"packages/ml/pytorch/torchtext/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchtext/README.md","title":"torchtext Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchtext%2FREADME.md -->","isInternal":false,"tokens":1053,"sizeBytes":3650},{"name":"README.md","path":"packages/ml/pytorch/torchvision/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/pytorch/torchvision/README.md","title":"torchvision Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fpytorch%2Ftorchvision%2FREADME.md -->","isInternal":false,"tokens":7340,"sizeBytes":19845},{"name":"README.md","path":"packages/ml/rapids/cudf/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/rapids/cudf/README.md","title":"cudf Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcudf%2FREADME.md -->","isInternal":false,"tokens":1648,"sizeBytes":5122},{"name":"README.md","path":"packages/ml/rapids/cuml/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/rapids/cuml/README.md","title":"cuml Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fcuml%2FREADME.md -->","isInternal":false,"tokens":1382,"sizeBytes":4364},{"name":"README.md","path":"packages/ml/rapids/raft/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/rapids/raft/README.md","title":"raft Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Frapids%2Fraft%2FREADME.md -->","isInternal":false,"tokens":827,"sizeBytes":2983},{"name":"README.md","path":"packages/ml/tensorflow/graphics/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/tensorflow/graphics/README.md","title":"graphics Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Fgraphics%2FREADME.md -->","isInternal":false,"tokens":1375,"sizeBytes":4689},{"name":"README.md","path":"packages/ml/tensorflow/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/tensorflow/README.md","title":"tensorflow Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2FREADME.md -->","isInternal":false,"tokens":2566,"sizeBytes":7995},{"name":"README.md","path":"packages/ml/tensorflow/tensorboard/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/tensorflow/tensorboard/README.md","title":"tensorboard Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftensorboard%2FREADME.md -->","isInternal":false,"tokens":768,"sizeBytes":2844},{"name":"README.md","path":"packages/ml/tensorflow/text/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/tensorflow/text/README.md","title":"text Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftensorflow%2Ftext%2FREADME.md -->","isInternal":false,"tokens":1315,"sizeBytes":4442},{"name":"README.md","path":"packages/ml/transformer-engine/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/transformer-engine/README.md","title":"transformer-engine Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftransformer-engine%2FREADME.md -->","isInternal":false,"tokens":1145,"sizeBytes":4043},{"name":"README.md","path":"packages/ml/triton/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/triton/README.md","title":"triton Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftriton%2FREADME.md -->","isInternal":false,"tokens":5782,"sizeBytes":16381},{"name":"README.md","path":"packages/ml/tritonserver/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/tritonserver/README.md","title":"tritonserver Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Ftritonserver%2FREADME.md -->","isInternal":false,"tokens":1572,"sizeBytes":4907},{"name":"README.md","path":"packages/ml/xlstm/mlstm_kernels/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/xlstm/mlstm_kernels/README.md","title":"mlstm_kernels Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fmlstm_kernels%2FREADME.md -->","isInternal":false,"tokens":1039,"sizeBytes":3611},{"name":"README.md","path":"packages/ml/xlstm/pltsm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/xlstm/pltsm/README.md","title":"pltsm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fpltsm%2FREADME.md -->","isInternal":false,"tokens":991,"sizeBytes":3473},{"name":"README.md","path":"packages/ml/xlstm/xlstm/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/ml/xlstm/xlstm/README.md","title":"xlstm Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fml%2Fxlstm%2Fxlstm%2FREADME.md -->","isInternal":false,"tokens":987,"sizeBytes":3472},{"name":"README.md","path":"packages/multimedia/decord/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/decord/README.md","title":"decord Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fdecord%2FREADME.md -->","isInternal":false,"tokens":1657,"sizeBytes":4857},{"name":"README.md","path":"packages/multimedia/ffmpeg/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/ffmpeg/README.md","title":"ffmpeg Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fffmpeg%2FREADME.md -->","isInternal":false,"tokens":4702,"sizeBytes":13043},{"name":"README.md","path":"packages/multimedia/gstreamer/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/gstreamer/README.md","title":"gstreamer Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fgstreamer%2FREADME.md -->","isInternal":false,"tokens":2595,"sizeBytes":7324},{"name":"README.md","path":"packages/multimedia/jetson-utils/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/jetson-utils/README.md","title":"jetson-utils Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fjetson-utils%2FREADME.md -->","isInternal":false,"tokens":2482,"sizeBytes":7259},{"name":"README.md","path":"packages/multimedia/libcom/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/libcom/README.md","title":"libcom Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Flibcom%2FREADME.md -->","isInternal":false,"tokens":961,"sizeBytes":3388},{"name":"README.md","path":"packages/multimedia/opengl/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/opengl/README.md","title":"opengl Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fopengl%2FREADME.md -->","isInternal":false,"tokens":4144,"sizeBytes":11516},{"name":"README.md","path":"packages/multimedia/pyav/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/pyav/README.md","title":"pyav Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fpyav%2FREADME.md -->","isInternal":false,"tokens":833,"sizeBytes":2988},{"name":"README.md","path":"packages/multimedia/sound-utils/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/sound-utils/README.md","title":"sound-utils Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fsound-utils%2FREADME.md -->","isInternal":false,"tokens":88,"sizeBytes":321},{"name":"README.md","path":"packages/multimedia/video-codec-sdk/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/video-codec-sdk/README.md","title":"video-codec-sdk Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvideo-codec-sdk%2FREADME.md -->","isInternal":false,"tokens":4673,"sizeBytes":13244},{"name":"README.md","path":"packages/multimedia/vulkan/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/multimedia/vulkan/README.md","title":"vulkan Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fmultimedia%2Fvulkan%2FREADME.md -->","isInternal":false,"tokens":4794,"sizeBytes":13551},{"name":"README.md","path":"packages/net/devpi/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/net/devpi/README.md","title":"devpi Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fdevpi%2FREADME.md -->","isInternal":false,"tokens":702,"sizeBytes":2587},{"name":"README.md","path":"packages/net/pypi.dev/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/net/pypi.dev/README.md","title":"pypi.dev Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fnet%2Fpypi.dev%2FREADME.md -->","isInternal":false,"tokens":105,"sizeBytes":436},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos_diffusion_renderer/nvdiffrast/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos_diffusion_renderer/nvdiffrast/README.md","title":"nvdiffrast Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2Fnvdiffrast%2FREADME.md -->","isInternal":false,"tokens":965,"sizeBytes":3339},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos_diffusion_renderer/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos_diffusion_renderer/README.md","title":"cosmos_diffusion_renderer Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos_diffusion_renderer%2FREADME.md -->","isInternal":false,"tokens":1139,"sizeBytes":4073},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos-policy/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos-policy/README.md","title":"cosmos-policy Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-policy%2FREADME.md -->","isInternal":false,"tokens":1091,"sizeBytes":3882},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos-predict2/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos-predict2/README.md","title":"cosmos-predict2 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-predict2%2FREADME.md -->","isInternal":false,"tokens":1091,"sizeBytes":3882},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos-reason1/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos-reason1/README.md","title":"cosmos-reason1 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-reason1%2FREADME.md -->","isInternal":false,"tokens":1223,"sizeBytes":4289},{"name":"README.md","path":"packages/physicalAI/cosmos/cosmos-transfer1/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/cosmos/cosmos-transfer1/README.md","title":"cosmos-transfer1 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fcosmos%2Fcosmos-transfer1%2FREADME.md -->","isInternal":false,"tokens":1083,"sizeBytes":3890},{"name":"README.md","path":"packages/physicalAI/isaac-ros/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/isaac-ros/README.md","title":"isaac-ros Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fisaac-ros%2FREADME.md -->","isInternal":false,"tokens":10957,"sizeBytes":33222},{"name":"README.md","path":"packages/physicalAI/lerobot/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/lerobot/README.md","title":"lerobot Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Flerobot%2FREADME.md -->","isInternal":false,"tokens":3318,"sizeBytes":10319},{"name":"README.md","path":"packages/physicalAI/opendronemap/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/opendronemap/README.md","title":"opendronemap Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopendronemap%2FREADME.md -->","isInternal":false,"tokens":1606,"sizeBytes":5401},{"name":"README.md","path":"packages/physicalAI/openpi/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/openpi/README.md","title":"openpi Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fopenpi%2FREADME.md -->","isInternal":false,"tokens":1116,"sizeBytes":3938},{"name":"README.md","path":"packages/physicalAI/protomotions/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/protomotions/README.md","title":"protomotions Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fprotomotions%2FREADME.md -->","isInternal":false,"tokens":1070,"sizeBytes":3748},{"name":"README.md","path":"packages/physicalAI/ros/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/ros/README.md","title":"ros Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fros%2FREADME.md -->","isInternal":false,"tokens":20166,"sizeBytes":50030},{"name":"README.md","path":"packages/physicalAI/sim/genesis/quadrants/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/genesis/quadrants/README.md","title":"quadrants Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fquadrants%2FREADME.md -->","isInternal":false,"tokens":901,"sizeBytes":3207},{"name":"Readme.md","path":"packages/physicalAI/sim/genesis/Readme.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/genesis/Readme.md","title":"genesis Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FReadme.md -->","isInternal":false,"tokens":1430,"sizeBytes":4714},{"name":"README.md","path":"packages/physicalAI/sim/genesis/splashSurf/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/genesis/splashSurf/README.md","title":"splashSurf Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2FsplashSurf%2FREADME.md -->","isInternal":false,"tokens":851,"sizeBytes":3072},{"name":"README.md","path":"packages/physicalAI/sim/genesis/vtk/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/genesis/vtk/README.md","title":"vtk Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fgenesis%2Fvtk%2FREADME.md -->","isInternal":false,"tokens":1186,"sizeBytes":4007},{"name":"README.md","path":"packages/physicalAI/sim/habitat-sim/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/habitat-sim/README.md","title":"habitat-sim Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fhabitat-sim%2FREADME.md -->","isInternal":false,"tokens":838,"sizeBytes":3003},{"name":"README.md","path":"packages/physicalAI/sim/isaac-sim/isaac-lab/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/isaac-sim/isaac-lab/README.md","title":"isaac-lab Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2Fisaac-lab%2FREADME.md -->","isInternal":false,"tokens":938,"sizeBytes":3282},{"name":"README.md","path":"packages/physicalAI/sim/isaac-sim/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/isaac-sim/README.md","title":"isaac-sim Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fisaac-sim%2FREADME.md -->","isInternal":false,"tokens":960,"sizeBytes":3333},{"name":"README.md","path":"packages/physicalAI/sim/mimicgen/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/mimicgen/README.md","title":"mimicgen Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmimicgen%2FREADME.md -->","isInternal":false,"tokens":1727,"sizeBytes":5128},{"name":"README.md","path":"packages/physicalAI/sim/mujoco/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/mujoco/README.md","title":"mujoco Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fmujoco%2FREADME.md -->","isInternal":false,"tokens":1589,"sizeBytes":4709},{"name":"README.md","path":"packages/physicalAI/sim/newton/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/newton/README.md","title":"newton Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fnewton%2FREADME.md -->","isInternal":false,"tokens":776,"sizeBytes":2813},{"name":"README.md","path":"packages/physicalAI/sim/physx/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/physx/README.md","title":"physx Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fphysx%2FREADME.md -->","isInternal":false,"tokens":776,"sizeBytes":2813},{"name":"README.md","path":"packages/physicalAI/sim/robogen/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/robogen/README.md","title":"robogen Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobogen%2FREADME.md -->","isInternal":false,"tokens":1022,"sizeBytes":3599},{"name":"README.md","path":"packages/physicalAI/sim/robomimic/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/robomimic/README.md","title":"robomimic Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobomimic%2FREADME.md -->","isInternal":false,"tokens":1753,"sizeBytes":5192},{"name":"README.md","path":"packages/physicalAI/sim/robosuite/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/robosuite/README.md","title":"robosuite Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Frobosuite%2FREADME.md -->","isInternal":false,"tokens":1701,"sizeBytes":5051},{"name":"README.md","path":"packages/physicalAI/sim/sapien/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/sim/sapien/README.md","title":"sapien Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fsim%2Fsapien%2FREADME.md -->","isInternal":false,"tokens":774,"sizeBytes":2818},{"name":"README.md","path":"packages/physicalAI/vla/crossformer/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/vla/crossformer/README.md","title":"crossformer Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fcrossformer%2FREADME.md -->","isInternal":false,"tokens":1262,"sizeBytes":4286},{"name":"README.md","path":"packages/physicalAI/vla/isaac-gr00t/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/vla/isaac-gr00t/README.md","title":"isaac-gr00t Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fisaac-gr00t%2FREADME.md -->","isInternal":false,"tokens":1471,"sizeBytes":4837},{"name":"README.md","path":"packages/physicalAI/vla/octo/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/vla/octo/README.md","title":"octo Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Focto%2FREADME.md -->","isInternal":false,"tokens":1256,"sizeBytes":4202},{"name":"README.md","path":"packages/physicalAI/vla/openvla/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/vla/openvla/README.md","title":"openvla Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Fopenvla%2FREADME.md -->","isInternal":false,"tokens":1906,"sizeBytes":6129},{"name":"README.md","path":"packages/physicalAI/vla/robopoint/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/physicalAI/vla/robopoint/README.md","title":"robopoint Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FphysicalAI%2Fvla%2Frobopoint%2FREADME.md -->","isInternal":false,"tokens":1465,"sizeBytes":4764},{"name":"README.md","path":"packages/rag/graphiti/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/rag/graphiti/README.md","title":"graphiti Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fgraphiti%2FREADME.md -->","isInternal":false,"tokens":739,"sizeBytes":2732},{"name":"README.md","path":"packages/rag/jetson-copilot/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/rag/jetson-copilot/README.md","title":"jetson-copilot Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fjetson-copilot%2FREADME.md -->","isInternal":false,"tokens":1924,"sizeBytes":6102},{"name":"README.md","path":"packages/rag/langchain/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/rag/langchain/README.md","title":"langchain Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Flangchain%2FREADME.md -->","isInternal":false,"tokens":2494,"sizeBytes":7414},{"name":"README.md","path":"packages/rag/llama-index/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/rag/llama-index/README.md","title":"llama-index Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fllama-index%2FREADME.md -->","isInternal":false,"tokens":2580,"sizeBytes":7996},{"name":"README.md","path":"packages/rag/memvid/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/rag/memvid/README.md","title":"memvid Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Frag%2Fmemvid%2FREADME.md -->","isInternal":false,"tokens":867,"sizeBytes":3110},{"name":"README.md","path":"packages/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/README.md","title":"packages Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2FREADME.md -->","isInternal":false,"tokens":8228,"sizeBytes":23693},{"name":"README.md","path":"packages/smart-home/dependencies/ciso8601/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/dependencies/ciso8601/README.md","title":"ciso8601 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fciso8601%2FREADME.md -->","isInternal":false,"tokens":795,"sizeBytes":2834},{"name":"README.md","path":"packages/smart-home/dependencies/psutil-home-assistant/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/dependencies/psutil-home-assistant/README.md","title":"psutil-home-assistant Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fdependencies%2Fpsutil-home-assistant%2FREADME.md -->","isInternal":false,"tokens":835,"sizeBytes":3053},{"name":"README.md","path":"packages/smart-home/homeassistant-base/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/homeassistant-base/README.md","title":"homeassistant-base Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-base%2FREADME.md -->","isInternal":false,"tokens":981,"sizeBytes":3536},{"name":"README.md","path":"packages/smart-home/homeassistant-core/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/homeassistant-core/README.md","title":"homeassistant-core Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fhomeassistant-core%2FREADME.md -->","isInternal":false,"tokens":3189,"sizeBytes":11224},{"name":"README.md","path":"packages/smart-home/wyoming/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/wyoming/README.md","title":"wyoming Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2FREADME.md -->","isInternal":false,"tokens":81,"sizeBytes":300},{"name":"README.md","path":"packages/smart-home/wyoming/wyoming-assist-microphone/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/wyoming/wyoming-assist-microphone/README.md","title":"wyoming-assist-microphone Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-assist-microphone%2FREADME.md -->","isInternal":false,"tokens":3338,"sizeBytes":11632},{"name":"README.md","path":"packages/smart-home/wyoming/wyoming-openwakeword/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/wyoming/wyoming-openwakeword/README.md","title":"wyoming-openwakeword Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-openwakeword%2FREADME.md -->","isInternal":false,"tokens":3229,"sizeBytes":11283},{"name":"README.md","path":"packages/smart-home/wyoming/wyoming-piper/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/wyoming/wyoming-piper/README.md","title":"wyoming-piper Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-piper%2FREADME.md -->","isInternal":false,"tokens":3161,"sizeBytes":10627},{"name":"README.md","path":"packages/smart-home/wyoming/wyoming-whisper/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/smart-home/wyoming/wyoming-whisper/README.md","title":"wyoming-whisper Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fsmart-home%2Fwyoming%2Fwyoming-whisper%2FREADME.md -->","isInternal":false,"tokens":2809,"sizeBytes":9684},{"name":"README.md","path":"packages/speech/audiocraft/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/audiocraft/README.md","title":"audiocraft Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Faudiocraft%2FREADME.md -->","isInternal":false,"tokens":1840,"sizeBytes":5663},{"name":"README.md","path":"packages/speech/chatterbox-tts/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/chatterbox-tts/README.md","title":"chatterbox-tts Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fchatterbox-tts%2FREADME.md -->","isInternal":false,"tokens":1561,"sizeBytes":6126},{"name":"README.md","path":"packages/speech/espeak/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/espeak/README.md","title":"espeak Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fespeak%2FREADME.md -->","isInternal":false,"tokens":1562,"sizeBytes":4594},{"name":"README.md","path":"packages/speech/faster-whisper/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/faster-whisper/README.md","title":"faster-whisper Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Ffaster-whisper%2FREADME.md -->","isInternal":false,"tokens":1320,"sizeBytes":4387},{"name":"README.md","path":"packages/speech/kokoro-tts/kokoro-tts-fastapi/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/kokoro-tts/kokoro-tts-fastapi/README.md","title":"kokoro-tts-fastapi Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-fastapi%2FREADME.md -->","isInternal":false,"tokens":1286,"sizeBytes":4311},{"name":"README.md","path":"packages/speech/kokoro-tts/kokoro-tts-hf/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/kokoro-tts/kokoro-tts-hf/README.md","title":"kokoro-tts-hf Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-hf%2FREADME.md -->","isInternal":false,"tokens":992,"sizeBytes":3398},{"name":"README.md","path":"packages/speech/kokoro-tts/kokoro-tts-onnx/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/kokoro-tts/kokoro-tts-onnx/README.md","title":"kokoro-tts-onnx Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fkokoro-tts%2Fkokoro-tts-onnx%2FREADME.md -->","isInternal":false,"tokens":918,"sizeBytes":3225},{"name":"README.md","path":"packages/speech/piper1-tts/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/piper1-tts/README.md","title":"piper1-tts Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fpiper1-tts%2FREADME.md -->","isInternal":false,"tokens":956,"sizeBytes":3356},{"name":"README.md","path":"packages/speech/riva-client/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/riva-client/README.md","title":"riva-client Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Friva-client%2FREADME.md -->","isInternal":false,"tokens":3560,"sizeBytes":10624},{"name":"README.md","path":"packages/speech/spark-tts/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/spark-tts/README.md","title":"spark-tts Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspark-tts%2FREADME.md -->","isInternal":false,"tokens":1458,"sizeBytes":5513},{"name":"README.md","path":"packages/speech/speaches/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/speaches/README.md","title":"speaches Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeaches%2FREADME.md -->","isInternal":false,"tokens":1615,"sizeBytes":5511},{"name":"README.md","path":"packages/speech/speech-dispatcher/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/speech-dispatcher/README.md","title":"speech-dispatcher Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fspeech-dispatcher%2FREADME.md -->","isInternal":false,"tokens":1555,"sizeBytes":5950},{"name":"README.md","path":"packages/speech/voice-pro/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/voice-pro/README.md","title":"voice-pro Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoice-pro%2FREADME.md -->","isInternal":false,"tokens":1158,"sizeBytes":4086},{"name":"README.md","path":"packages/speech/voicecraft/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/voicecraft/README.md","title":"voicecraft Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fvoicecraft%2FREADME.md -->","isInternal":false,"tokens":1564,"sizeBytes":5166},{"name":"README.md","path":"packages/speech/whisper_trt/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/whisper_trt/README.md","title":"whisper_trt Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper_trt%2FREADME.md -->","isInternal":false,"tokens":1981,"sizeBytes":5795},{"name":"README.md","path":"packages/speech/whisper/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/whisper/README.md","title":"whisper Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisper%2FREADME.md -->","isInternal":false,"tokens":1811,"sizeBytes":5986},{"name":"README.md","path":"packages/speech/whisperx/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/whisperx/README.md","title":"whisperx Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fwhisperx%2FREADME.md -->","isInternal":false,"tokens":1397,"sizeBytes":4527},{"name":"README.md","path":"packages/speech/xtts/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/speech/xtts/README.md","title":"xtts Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fspeech%2Fxtts%2FREADME.md -->","isInternal":false,"tokens":1269,"sizeBytes":4195},{"name":"README.md","path":"packages/vectordb/faiss_lite/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vectordb/faiss_lite/README.md","title":"faiss_lite Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss_lite%2FREADME.md -->","isInternal":false,"tokens":1614,"sizeBytes":4775},{"name":"README.md","path":"packages/vectordb/faiss/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vectordb/faiss/README.md","title":"faiss Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Ffaiss%2FREADME.md -->","isInternal":false,"tokens":2204,"sizeBytes":6205},{"name":"README.md","path":"packages/vectordb/nanodb/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vectordb/nanodb/README.md","title":"nanodb Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvectordb%2Fnanodb%2FREADME.md -->","isInternal":false,"tokens":3921,"sizeBytes":12311},{"name":"README.md","path":"packages/vit/clip_trt/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/clip_trt/README.md","title":"clip_trt Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fclip_trt%2FREADME.md -->","isInternal":false,"tokens":1708,"sizeBytes":5047},{"name":"README.md","path":"packages/vit/efficientvit/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/efficientvit/README.md","title":"efficientvit Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fefficientvit%2FREADME.md -->","isInternal":false,"tokens":1526,"sizeBytes":4965},{"name":"README.md","path":"packages/vit/nanoowl/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/nanoowl/README.md","title":"nanoowl Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanoowl%2FREADME.md -->","isInternal":false,"tokens":2098,"sizeBytes":6499},{"name":"README.md","path":"packages/vit/nanosam/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/nanosam/README.md","title":"nanosam Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fnanosam%2FREADME.md -->","isInternal":false,"tokens":1700,"sizeBytes":5297},{"name":"README.md","path":"packages/vit/sam/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/sam/README.md","title":"sam Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsam%2FREADME.md -->","isInternal":false,"tokens":1937,"sizeBytes":6364},{"name":"README.md","path":"packages/vit/sapiens/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/sapiens/README.md","title":"sapiens Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Fsapiens%2FREADME.md -->","isInternal":false,"tokens":963,"sizeBytes":3438},{"name":"README.md","path":"packages/vit/tam/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vit/tam/README.md","title":"tam Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvit%2Ftam%2FREADME.md -->","isInternal":false,"tokens":1624,"sizeBytes":5091},{"name":"README.md","path":"packages/vlm/lita/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/lita/README.md","title":"lita Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Flita%2FREADME.md -->","isInternal":false,"tokens":1395,"sizeBytes":4553},{"name":"README.md","path":"packages/vlm/llama-vision/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/llama-vision/README.md","title":"llama-vision Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllama-vision%2FREADME.md -->","isInternal":false,"tokens":1227,"sizeBytes":4113},{"name":"README.md","path":"packages/vlm/llava/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/llava/README.md","title":"llava Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fllava%2FREADME.md -->","isInternal":false,"tokens":2151,"sizeBytes":6738},{"name":"README.md","path":"packages/vlm/minigpt4/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/minigpt4/README.md","title":"minigpt4 Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fminigpt4%2FREADME.md -->","isInternal":false,"tokens":1911,"sizeBytes":5828},{"name":"README.md","path":"packages/vlm/prismatic/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/prismatic/README.md","title":"prismatic Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fprismatic%2FREADME.md -->","isInternal":false,"tokens":954,"sizeBytes":3429},{"name":"README.md","path":"packages/vlm/videollama/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/videollama/README.md","title":"videollama Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvideollama%2FREADME.md -->","isInternal":false,"tokens":1063,"sizeBytes":3765},{"name":"README.md","path":"packages/vlm/vila-microservice/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/vila-microservice/README.md","title":"vila-microservice Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila-microservice%2FREADME.md -->","isInternal":false,"tokens":1142,"sizeBytes":3767},{"name":"README.md","path":"packages/vlm/vila/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/vila/README.md","title":"vila Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fvila%2FREADME.md -->","isInternal":false,"tokens":2264,"sizeBytes":6544},{"name":"README.md","path":"packages/vlm/xtuner/README.md","rawUrl":"https://raw.githubusercontent.com/dusty-nv/jetson-containers/HEAD/packages/vlm/xtuner/README.md","title":"xtuner Documentation","category":"plugin-manifest","format":"markdown","content":"<!-- Skill available on-demand via /api/skills/dusty-nv/jetson-containers?path=packages%2Fvlm%2Fxtuner%2FREADME.md -->","isInternal":false,"tokens":1255,"sizeBytes":4219}],"systemPromptSnippet":"<yakaai_skills repo=\"dusty-nv/jetson-containers\">\n<!-- File: AGENTS.md (Tokens: ~2402 | Category: root-instruction) -->\n# CLAUDE.md\n\nThis file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.\n\n## Overview\n\njetson-containers is a modular container build system for NVIDIA Jetson devices. It discovers packages, resolves their dependencies, chains their Dockerfiles together, and runs `docker build` to produce multi-stage images optimized for specific L4T/JetPack/CUDA versions.\n\n## Commands\n\n### Build & Run\n```bash\n# Build one or more packages (chains Dockerfiles in dependency order)\njetson-containers build <package> [<package2> ...]\n# or equivalently:\n./build.sh <package> [<package2> ...]\n\n# Find and run a compatible container image\njetson-containers run $(autotag <package>) [command]\n# or equivalently:\n./run.sh $(autotag <package>) [command]\n\n# Find the best matching image tag (local → DockerHub → build)\nautotag <package>\n\n# Useful build flags\njetson-containers build --simulate <package> # dry-run: print commands without building\njetson-containers build --multiple pytorch tensorflow # build each package as a separate image\njetson-containers build --list-packages # list all discovered packages\njetson-containers build --show-packages <package> # show resolved metadata for a package\njetson-containers build --base=xyz:latest <package> # start the build chain from an existing image\njetson-containers build --push=myuser <package> # push to a registry after building\n\n# Check for upstream version upgrades (--apply to write changes)\njetson-containers upgrade [<package> ...] [--apply]\n```\n\n### Code Style\n```bash\n# Install pre-commit hooks (one-time setup)\npip install -r requirements.txt\npre-commit install\n\n# Run linting/formatting checks manually\npre-commit run --all-files\n```\n\nPre-commit runs **Black** (formatter, line-length=88) and **Flake8** (linter) only on `jetson_containers/` and `test_precommit.py`. `packages/` Dockerfiles and `config.py` files are **not** linted by pre-commit. Note: `skip-string-normalization = true` in `pyproject.toml`.\n\n### Tests\n```bash\n# Run a package config unit test directly (uses stdlib unittest, no pytest needed)\npython3 tests/verify_triton_config.py\n```\n\n### Installation\n```bash\nbash install.sh # sets up venv, installs deps, symlinks CLI tools to /usr/local/bin\n```\n\n## Architecture\n\n### Package System\n\nEvery buildable unit is a **package**. Packages live under `packages/<category>/<name>/` and define metadata via one of (in priority order):\n\n1. **YAML header in Dockerfile** — metadata between `#---` markers at the top\n2. **`config.yaml` / `config.yml`** — static YAML config file\n3. **`config.json`** — meta-container (no Dockerfile, just dependency composition)\n4. **`config.py`** — dynamic config, returns a list of package dicts at build time\n\nKey metadata fields: `name`, `alias`, `depends` (build-order deps), `requires` (version constraints on L4T/CUDA/Python), `test`, `build_args`, `dockerfile`.\n\n### Build Pipeline\n\n1. **`jetson_containers/packages.py`** — scans `packages/` recursively, runs `config.py` scripts, resolves the dependency graph, filters by L4T version compatibility, and exposes a custom Python meta path finder so packages can `from packages.ml.pytorch.version import ...`\n2. **`jetson_containers/build.py`** — CLI entry point; parses args and dispatches\n3. **`jetson_containers/container.py`** — core build logic: stitches multiple package Dockerfiles into one combined `Dockerfile`, invokes `docker build`, optionally runs per-package tests, and optionally pushes to a registry\n4. **`jetson_containers/l4t_version.py`** — detects architecture (tegra-aarch64, aarch64, x86_64) and reads L4T/JetPack/CUDA/GPU-arch info from `/etc/nv_tegra_release`; all version variables (e.g. `CUDA_VERSION`, `PYTORCH_VERSION`) flow from here into build args\n\n### Dynamic Package Generation\n\n`config.py` files execute at build time and can return multiple package dicts to produce version variants. For example, PyTorch's `config.py` generates `pytorch:2.8`, `pytorch:2.8-all`, `pytorch:2.8-builder`, etc. Aliases (`torch` → `pytorch:2.8`) are resolved during package discovery.\n\nWhen a package's version selection logic is complex, put it in a separate `version.py` alongside `config.py`. This lets other packages cross-import it:\n```python\nfrom packages.ml.pytorch.version import PYTORCH_VERSION\n```\nThe `_PackagesFinder` meta path finder in `packages.py` resolves these imports against the `packages/` tree without requiring `__init__.py` files.\n\n### Dockerfile Conventions\n\nUse `uv pip install` (not plain `pip install`) inside Dockerfiles — `uv` is pre-installed in the base images and is significantly faster. Every Dockerfile must start with:\n```dockerfile\nARG BASE_IMAGE\nFROM ${BASE_IMAGE}\n```\n\n### Version Overrides\n\nCopy `.env.default` → `.env` to override `L4T_VERSION`, `CUDA_VERSION`, `LSB_RELEASE`, `CUDA_ARCH`, local PyPI/APT mirror URLs, and SCP/webhook credentials without modifying tracked files.\n\n### Creating a New Package\n\n1. Create `packages/<category>/<name>/Dockerfile` with a `#---` YAML header (or a `config.yaml`/`config.py`).\n2. Declare `depends` on upstream packages (e.g., `pytorch`, `cuda`).\n3. Optionally add `test.py` or `test.sh` for post-build validation.\n4. Build: `jetson-containers build <name>`\n\n## Security Features\n\n### Multi-Stage Builds\n\nThe build system is multi-stage aware. Packages can use `FROM ... AS <alias>` for intermediate stages (e.g. a compiler/builder stage) without those stages polluting the final image. The build system tracks stage boundaries and only injects BuildKit features (ccache env vars, `--device` mounts, ccache `--mount` flags) into the **final** stage (the `FROM ${BASE_IMAGE}` stage). Intermediate stages are left untouched.\n\nDeclare multi-stage packages in config metadata:\n```yaml\n#---\n# name: mypackage\n# multistage: true\n# depends: [cuda]\n#---\nARG BASE_IMAGE\nFROM gcc:12 AS builder\nRUN make install\n\nFROM ${BASE_IMAGE}\nCOPY --from=builder /usr/local /usr/local\nRUN ldconfig\n```\n\nThe final stage **must** use `ARG BASE_IMAGE` / `FROM ${BASE_IMAGE}` so it chains correctly in the dependency build pipeline.\n\n### Non-Root User\n\n**At build time** — pass identity via env vars; packages can create and switch to the declared user:\n```bash\nexport CONTAINER_USER=jetson\nexport CONTAINER_UID=1000\nexport CONTAINER_GID=1000\njetson-containers build mypackage\n```\n\n`CONTAINER_USER`, `CONTAINER_UID`, `CONTAINER_GID` are forwarded as `--build-arg` automatically. Packages consume them like:\n```dockerfile\nARG CONTAINER_USER=root\nARG CONTAINER_UID=0\nARG CONTAINER_GID=0\nRUN if [ \"$CONTAINER_UID\" != \"0\" ]; then \\\n groupadd -g $CONTAINER_GID $CONTAINER_USER && \\\n useradd -u $CONTAINER_UID -g $CONTAINER_GID -m $CONTAINER_USER; \\\n fi\nUSER $CONTAINER_USER\n```\n\nDeclare the supported user in package metadata (`run_user` is informational for tooling):\n```yaml\n# run_user: jetson\n```\n\n**At runtime** — set `DOCKER_USER` to run the container as a non-root identity:\n```bash\nDOCKER_USER=1000:1000 ./run.sh myimage\n# or by name (the user must exist in the image):\nDOCKER_USER=jetson ./run.sh myimage\n```\nX11 forwarding (`xhost`) is automatically updated to grant the declared user display access.\n\nFor GPU access, `run.sh` also discovers the numeric GIDs that own the NVIDIA/Tegra device nodes (`/dev/nvhost*`, `/dev/nvmap`, `/dev/nvgpu/*`, `/dev/dri/render*`) and adds them via `--group-add <gid>`. Without this, a non-root process can't open the GPU nodes and CUDA fails with error 801 (`cudaGetDeviceCount` \"operation not supported\"). Numeric GIDs are used because the host's `render` GID usually doesn't match the container's. See [`docs/run.md`](/docs/run.md#gpu-access-as-non-root-cuda-error-801).\n\n### Secrets\n\n**Build-time secrets** (never baked into image layers) — declare secrets in package metadata:\n```yaml\n#---\n# name: mypackage\n# secrets: [huggingface_token]\n#---\n```\n\nPoint each secret at a file on the host via `JETSON_SECRET_<NAME>`:\n```bash\nexport JETSON_SECRET_HUGGINGFACE_TOKEN=~/.hf_token\njetson-containers build mypackage\n```\n\nSecrets are mounted as `/run/secrets/<name>` inside `RUN` steps only (BuildKit `--secret`). Use them in the Dockerfile:\n```dockerfile\nRUN --mount=type=secret,id=huggingface_token \\\n HF_TOKEN=$(cat /run/secrets/huggingface_token) \\\n python3 download_model.py\n```\n\nIf the env var is missing, a warning is printed and the build continues without that secret.\n\n**Runtime secrets** — prefer file-backed tokens over env vars (values in env vars are visible in `docker inspect`):\n```bash\n# Preferred: mount token file into /run/secrets/\nHUGGINGFACE_TOKEN_FILE=~/.hf_token ./run.sh myimage\n\n# Fallback (token visible in docker inspect):\nHUGGINGFACE_TOKEN=hf_xxx ./run.sh myimage\n```\n\n`HUGGINGFACE_TOKEN_FILE` mounts the file read-only at `/run/secrets/huggingface_token` and sets `HF_TOKEN_FILE` inside the container.\n\n### Key Supporting Modules\n\n- `jetson_containers/tag.py` — image tagging conventions\n- `jetson_containers/docs.py` — auto-generates per-package docs from metadata\n- `jetson_containers/ci.py` — CI/CD helpers\n- `jetson_containers/network.py` / `webhook.py` — GitHub API integration and build notifications\n- `jetson_containers/logging.py` — colored terminal output used throughout\n- `jetson_containers/upgrade.py` — checks packages against upstream (PyPI/GitHub) for newer versions; powers `jetson-containers upgrade`\n- `jetson_containers/db.py` — syncs DockerHub/GitHub metadata for registry introspection\n\n<!-- File: CLAUDE.md (Tokens: ~5 | Category: claude-rule) -->\n@AGENTS.md\n\n<!-- File: .claude/skills/jetson-pr.md (Tokens: ~751 | Category: anthropic-skill) -->\n# Create a jetson-containers Pull Request\n\nHelp the user create a well-formed PR that follows this repository's conventions.\n\n## Repository PR conventions\n\n- **Target branch**: PRs go to `dev`, not `master`. The CI workflow `pr-on-dev.yml` runs self-hosted Jetson builds against `dev`.\n- **Scope**: each PR should touch only the packages it claims to change. Keep unrelated cleanup out.\n- **Commit style**: short, lowercase, imperative subject. No period. Examples from git log: `fix`, `cuda 13.2 needs newer drivers to run in Orin`, `tensorrt compatible orin`, `Update ffmpeg version from 8.1 to 8.0.1`.\n- **No force-push** to shared branches.\n\n## Step-by-step workflow\n\n1. **Understand the changes**: run `git diff master...HEAD` and `git log master..HEAD --oneline` to see exactly what is changing. Read the changed files if needed to understand the intent.\n\n2. **Identify affected packages**: map every changed file back to its package directory (`packages/<category>/<name>/`). List them — this drives the PR title and body.\n\n3. **Stage and commit** (if there are uncommitted changes):\n - Stage only the files related to this PR.\n - Write a commit message following the style above. No trailing period.\n - Do not skip hooks (`--no-verify`).\n\n4. **Create or switch to a feature branch** if on `master`:\n ```\n git checkout -b <short-descriptive-name>\n ```\n Branch name: lowercase, hyphenated, describes the change (e.g. `upgrade-pytorch-2.9`, `add-sglang-0.4`, `fix-tensorrt-orin`).\n\n5. **Push and open the PR**:\n ```\n gh pr create --base dev --title \"<title>\" --body \"<body>\"\n ```\n\n6. **PR title format**: `<verb> <package(s)> <version or brief description>`\n - Version bumps: `upgrade pytorch to 2.9`\n - New packages: `add sglang 0.4`\n - Bug fixes: `fix tensorrt build on Orin`\n - Multi-package: list up to 3, then `and N others`\n\n7. **PR body template**:\n ```\n ## Changes\n - <package>: <what changed and why>\n\n ## Packages affected\n - `packages/<category>/<name>/`\n\n ## Test plan\n - [ ] `jetson-containers build <package>` on Orin (JP6)\n - [ ] Add any version-specific test steps here\n\n ## Notes\n <Any JetPack version constraints, driver requirements, or upstream release links>\n ```\n\n8. **Return the PR URL** when done.\n\n## What good PRs look like in this repo\n\n- Version upgrades include the new version entry in `config.py` **and** update `version.py` if it exists.\n- New packages include a `Dockerfile`, metadata (YAML header or `config.yaml`), and at minimum a stub `test.py` or `test.sh`.\n- Dependency bumps update `requires` constraints if the new version drops support for an older JetPack.\n- PRs do not mix version bumps with unrelated refactors.\n\n## Before opening the PR, verify\n\n- `git diff --stat dev...HEAD` shows only the intended files.\n- No `.env`, secrets, or build artifacts are staged.\n- The branch name and title are consistent.\n\n<!-- File: .claude/skills/upgrade-package.md (Tokens: ~717 | Category: anthropic-skill) -->\n# Upgrade Package Version\n\nHelp the user upgrade the version of one or more packages in this jetson-containers repository.\n\n## What the user will provide\n\nThe user will tell you the package name and the new version. If they only give you the package name, ask for the target version before proceeding.\n\n## How version definitions work\n\nEach package under `packages/<category>/<name>/` defines its version in one of these ways:\n\n1. **`version.py`** — a standalone file that exports a version constant, e.g. `PYTORCH_VERSION = Version('2.8')`. This is the canonical version used by `config.py`.\n2. **`config.py`** — calls a helper function (like `pytorch_pip(version, requires=...)`) with explicit version strings in a `package = [...]` list. Adding a new version = adding a new line to this list.\n3. **`config.yaml` / YAML header in Dockerfile** — static `build_args` with version pinned directly.\n4. **`Dockerfile` `ARG` lines** — `ARG SOME_VERSION=x.y.z`.\n\n## Step-by-step workflow\n\n1. **Locate the package directory**: search under `packages/` for the package name.\n2. **Read all config files** in that directory (`config.py`, `config.yaml`, `version.py`, `Dockerfile`, `Dockerfile.*`).\n3. **Identify where the version is defined** (see above). Show the user the current version(s) before making changes.\n4. **Apply the upgrade**:\n - If there is a `version.py`, update the version constant there first; `config.py` will pick it up automatically.\n - If `config.py` uses a list of `pytorch_pip(version, ...)` calls, **add** a new entry for the new version (do not remove old ones — the repo keeps historical versions for different JetPack targets). Set the new version as the latest by updating the `PYTORCH_VERSION`-like constant in `version.py`.\n - If the version is in a YAML `build_args` or a Dockerfile `ARG`, do a targeted edit of that value.\n5. **Check for cascading dependencies**: grep for the old version string across other config files in the same package directory. Common locations: `Dockerfile` `ARG` defaults, `build_args` dicts, `requires` strings, download URLs, wheel filenames.\n6. **Update download artifacts if needed**: if the Dockerfile fetches a wheel, tarball, or pre-built binary, update the URL/filename to match the new version. Check the upstream source (PyPI, NVIDIA, GitHub releases) if the user hasn't provided a URL.\n7. **Show a diff summary** of every file you changed before finishing.\n\n## Things to preserve\n\n- Do **not** delete existing version entries in `config.py` lists — old JetPack targets still need them.\n- Keep the `requires` constraints on old entries unchanged.\n- Do not change `alias` lists for older versions.\n- Follow the existing code style (no reformatting unrelated code, no added comments).\n\n## When to ask vs. proceed\n\n- Ask for the target version if not given.\n- Ask about JetPack `requires` constraint only if the new version introduces a minimum JetPack requirement that differs from the previous latest entry.\n- Proceed with everything else silently and show the diff at the end.\n\n<!-- File: packages/attention/block-sparse-attention/README.md (Tokens: ~1021 | Category: plugin-manifest) -->\n# block-sparse-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`block_sparse_attn:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `block_sparse_attn` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`xattention:0.0.1`](/packages/attention/xattention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag block-sparse-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host block-sparse-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag block-sparse-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag block-sparse-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build block-sparse-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/flash-attention/README.md (Tokens: ~4368 | Category: plugin-manifest) -->\n# flash-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flash-attention:2.5.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.6.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.7.4.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.0.post2`** | |\n| :-- | :-- |\n|    Aliases | `flash-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flash-attention:2.8.1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/flashinfer/README.md (Tokens: ~2975 | Category: plugin-manifest) -->\n# flash-infer\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flashinfer:0.2.1.post2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:0.8.4`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.2.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.6.post1`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`mlc:0.20.0`](/packages/llm/mlc) [`sglang:0.4.6`](/packages/llm/sglang) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.7`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`flashinfer:0.2.8`** | |\n| :-- | :-- |\n|    Aliases | `flashinfer` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) [`triton`](/packages/ml/triton) |\n|    Dependants | [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`llama-factory`](/packages/llm/llama-factory) [`local_llm`](/packages/llm/local_llm) [`mlc:0.21.0`](/packages/llm/mlc) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flash-infer)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flash-infer:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flash-infer)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flash-infer) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flash-infer\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/flexprefill/README.md (Tokens: ~1326 | Category: plugin-manifest) -->\n# flexprefill\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`flexprefill:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `flexprefill` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`llama-factory`](/packages/llm/llama-factory) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`sdnext`](/packages/diffusion/sdnext) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag flexprefill)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host flexprefill:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag flexprefill)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag flexprefill) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build flexprefill\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/huggingface_kernels/README.md (Tokens: ~968 | Category: plugin-manifest) -->\n# huggingface_kernels\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`huggingface_kernels:0.7.0`** | |\n| :-- | :-- |\n|    Aliases | `huggingface_kernels` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`huggingface_hub`](/packages/llm/huggingface_hub) [`torchvision`](/packages/pytorch/torchvision) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag huggingface_kernels)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host huggingface_kernels:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag huggingface_kernels)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag huggingface_kernels) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build huggingface_kernels\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/log-linear-attention/README.md (Tokens: ~976 | Category: plugin-manifest) -->\n# log-linear-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`log-linear-attention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `log-linear-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`ninja`](/packages/build/ninja) [`torchaudio`](/packages/pytorch/torchaudio) [`causalconv1d`](/packages/ml/mamba/causalconv1d) [`mamba`](/packages/ml/mamba/mamba) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag log-linear-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host log-linear-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag log-linear-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag log-linear-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build log-linear-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/ParaAttention/README.md (Tokens: ~1126 | Category: plugin-manifest) -->\n# ParaAttention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nhttps://github.com/chengzeyi/ParaAttention\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`paraattention:0.4.0`** | |\n| :-- | :-- |\n|    Aliases | `paraattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag ParaAttention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host ParaAttention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag ParaAttention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag ParaAttention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build ParaAttention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/radial-attention/README.md (Tokens: ~929 | Category: plugin-manifest) -->\n# radial-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`radial-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `radial-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`flashinfer`](/packages/attention/flash-infer) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag radial-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host radial-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag radial-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag radial-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build radial-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/sage-attention/README.md (Tokens: ~1135 | Category: plugin-manifest) -->\n# sage-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sage-attention:3.0.0`** | |\n| :-- | :-- |\n|    Aliases | `sage-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`framepack`](/packages/diffusion/framepack) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sage-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sage-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sage-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sage-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sage-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/sparge-attention/README.md (Tokens: ~1154 | Category: plugin-manifest) -->\n# sparge-attention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparge-attention:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparge-attention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`comfyui`](/packages/diffusion/comfyui) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`lobechat`](/packages/llm/lobe_chat) [`sdnext`](/packages/diffusion/sdnext) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/thu-ml/SpargeAttn.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparge-attention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparge-attention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparge-attention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparge-attention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparge-attention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/xattention/README.md (Tokens: ~1056 | Category: plugin-manifest) -->\n# xattention\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xattention:0.0.1`** | |\n| :-- | :-- |\n|    Aliases | `xattention` |\n|    Requires | `L4T ['>=35']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`block_sparse_attn`](/packages/attention/block-sparse-attention) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xattention)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host xattention:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xattention)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xattention) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xattention\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/attention/xformers/README.md (Tokens: ~3796 | Category: plugin-manifest) -->\n# xformers\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`xformers:0.0.32`** | |\n| :-- | :-- |\n|    Aliases | `xformers` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dependants | [`apex:0.1`](/packages/pytorch/apex) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n| **`xformers:0.0.33`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`triton`](/packages/ml/triton) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/facebookresearch/xformers |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/xformers:r35.2.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-06` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.3.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2023-12-14` | `arm64` | `5.8GB` |\n|   [`dustynv/xformers:r35.4.1`](https://hub.docker.com/r/dustynv/xformers/tags) | `2024-01-09` | `arm64` | `5.9GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag xformers)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/xformers:r35.4.1\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/xformers:r35.4.1\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag xformers)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag xformers) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build xformers\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/code/jupyterlab/README.md (Tokens: ~3020 | Category: plugin-manifest) -->\n# jupyterlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\nRunning the `jupyerlab` container will automatically start a JupyterLab server in the background on port 8888, with default login password `nvidia`. The JupyterLab server logs will be saved to `/data/logs/jupyter.log` if you need to inspect them (this location is automatically mounted under your `jetson-containers/data` directory)\n\nTo change the default settings, you can set the `$JUPYTER_ROOT`, `$JUPYTER_PORT`, `$JUPYTER_PASSWORD`, and `$JUPYTER_LOG` environment variables when starting the container like so:\n\n```bash\njetson-containers run \\\n --env JUPYTER_ROOT=/home/user \\\n --env JUPYTER_PORT=8000 \\\n --env JUPYTER_PASSWORD=password \\\n --env JUPYTER_LOGS=/dev/null \\\n $(autotag jupyterlab)\n```\n\nThe [`/start_jupyter`](./start_jupyter) script is the default CMD that the container runs when it starts - however, if you don't want the JupyterLab server started by default, you can either add a different CMD in your own Dockerfile, or override it at startup:\n\n```bash\n# skip straight to the terminal instead of starting JupyterLab first\njetson-containers run /bin/bash\n```\n\nYou can then still manually run the [`/start_jupyter`](./start_jupyter) script later when desired.\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`jupyterlab:latest`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:latest-myst`** | |\n| :-- | :-- |\n|    Aliases | `jupyterlab:myst` |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:latest`](/packages/code/jupyterlab) |\n|    Dependants | [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) |\n|    Dependants | [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n| **`jupyterlab:4.2.0-myst`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`tensorrt`](/packages/cuda/tensorrt) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`pycuda`](/packages/cuda/pycuda) [`rust`](/packages/build/rust) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) |\n|    Dockerfile | [`Dockerfile.myst`](Dockerfile.myst) |\n|    Notes | will autostart Jupyter server on port 8888 unless container entry CMD is overridden |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/jupyterlab:r32.7.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.7GB` |\n|   [`dustynv/jupyterlab:r35.2.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-12-06` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r35.3.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `5.4GB` |\n|   [`dustynv/jupyterlab:r35.4.1`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2023-10-07` | `arm64` | `5.3GB` |\n|   [`dustynv/jupyterlab:r36.2.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2024-03-07` | `arm64` | `0.6GB` |\n|   [`dustynv/jupyterlab:r36.4.0`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-10` | `arm64` | `5.8GB` |\n|   [`dustynv/jupyterlab:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/jupyterlab/tags) | `2025-03-03` | `arm64` | `5.1GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag jupyterlab)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/jupyterlab:r36.4.0\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/jupyterlab:r36.4.0\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag jupyterlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag jupyterlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build jupyterlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/code/vscode/README.md (Tokens: ~1526 | Category: plugin-manifest) -->\n# vscode\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`vscode`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:cuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`tensorrt`](/packages/cuda/tensorrt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:torch`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`triton`](/packages/ml/triton) [`torchao`](/packages/pytorch/torchao) [`tensorrt`](/packages/cuda/tensorrt) [`torch2trt`](/packages/pytorch/torch2trt) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n| **`vscode:transformers`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`triton`](/packages/ml/triton) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`bitsandbytes`](/packages/llm/bitsandbytes) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | installs portable build of VSCode inside container |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag vscode)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host vscode:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag vscode)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag vscode) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build vscode\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cuda/cuda-python/README.md (Tokens: ~5141 | Category: plugin-manifest) -->\n# cuda-python\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda-python:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda-python` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cuml`](/packages/ml/rapids/cuml) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss_lite`](/packages/vectordb/faiss_lite) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index:samples`](/packages/rag/llama-index) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`raft`](/packages/ml/rapids/raft) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`tam`](/packages/vit/tam) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.8`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu128`](/packages/cuda/cuda) [`cuda:12.8`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:12.9`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu129`](/packages/cuda/cuda) [`cuda:12.9`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`cuda-python:13.0`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu130`](/packages/cuda/cuda) [`cuda:13.0`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/cuda-python:builder-r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-26` | `arm64` | `5.1GB` |\n|   [`dustynv/cuda-python:builder-r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2024-03-11` | `arm64` | `3.6GB` |\n|   [`dustynv/cuda-python:r35.2.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.3.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r35.4.1`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/cuda-python:r36.2.0`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/cuda-python:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/cuda-python/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cuda-python)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/cuda-python:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/cuda-python:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cuda-python)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cuda-python) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cuda-python\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cuda/cuda/README.md (Tokens: ~19059 | Category: plugin-manifest) -->\n# cuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cuda:12.2`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/cuda:12.2-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-05, 3.4GB)`<br>[`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n\n| **`pip_cache:cu122`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.2`](/packages/cuda/cuda-python) [`cuda:12.2`](/packages/cuda/cuda) [`cuda:12.2-samples`](/packages/cuda/cuda) [`cudnn:8.9`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.2-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu122`](/packages/cuda/cuda) [`cuda:12.2`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Images | [`dustynv/cuda:12.2-samples-r36.2.0`](https://hub.docker.com/r/dustynv/cuda/tags) `(2023-12-07, 4.8GB)` |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu124`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`cuda-python:12.4`](/packages/cuda/cuda-python) [`cuda:12.4`](/packages/cuda/cuda) [`cuda:12.4-samples`](/packages/cuda/cuda) [`cudnn:9.0`](/packages/cuda/cudnn) |\n|    Dockerfile | [`Dockerfile.pip`](Dockerfile.pip) |\n\n| **`cuda:12.4-samples`** | |\n| :-- | :-- |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu124`](/packages/cuda/cuda) [`cuda:12.4`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) |\n|    Dockerfile | [`Dockerfile.samples`](Dockerfile.samples) |\n|    Notes | CUDA samples from https://github.com/NVIDIA/cuda-samples installed under /opt/cuda-samples |\n\n| **`cuda:12.6`** | |\n| :-- | :-- |\n|    Aliases | `cuda` |\n|    Requires | `L4T ['==36.*', 'aarch64']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`clip_trt`](/packages/vit/clip_trt) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`protomotions:2.5.0`](/packages/robots/protomotions) [`pyav`](/packages/multimedia/pyav) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/packages/speech/whisperx) [`wyoming-piper:1.6.2`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-piper:master`](/packages/smart-home/wyoming/wyoming-piper) [`wyoming-whisper:2.5.0`](/packages/smart-home/wyoming/wyoming-whisper) [`wyoming-whisper:master`](/packages/smart-home/wyoming/wyoming-whisper) [`xattention:0.0.1`](/packages/attention/xattention) [`xformers:0.0.32`](/packages/attention/xformers) [`xformers:0.0.33`](/packages/attention/xformers) [`xgrammar:0.1.15`](/packages/llm/xgrammar) [`xgrammar:0.1.18`](/packages/llm/xgrammar) [`xgrammar:0.1.19`](/packages/llm/xgrammar) [`xgrammar:0.1.20`](/packages/llm/xgrammar) [`xgrammar:0.1.21`](/packages/llm/xgrammar) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtts`](/packages/speech/xtts) [`xtuner`](/packages/vlm/xtuner) [`zed:5.0`](/packages/hw/zed) [`zed:5.0-humble`](/packages/hw/zed) [`zed:5.0-jazzy`](/packages/hw/zed) [`zigma:1.0`](/packages/ml/mamba/zigma) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n| **`pip_cache:cu126`** | |\n| :-- | :-- |\n|    Aliases | `pip_cache` `pip-cache` |\n|    Requires | `L4T ['==36.*']` |\n|    Dependencies | |\n|    Dependants | [`3d_diffusion_policy`](/packages/diffusion/3d_diffusion_policy) [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) [`4k4d:0.0.0`](/packages/3d/gaussian_splatting/4k4d) [`ai-toolkit`](/packages/diffusion/ai-toolkit) [`aim`](/packages/ml/aim) [`apex:0.1`](/packages/pytorch/apex) [`arrow:12.0.1`](/packages/numeric/arrow) [`arrow:14.0.1`](/packages/numeric/arrow) [`arrow:19.0.1`](/packages/numeric/arrow) [`arrow:20.0.0`](/packages/numeric/arrow) [`arrow:5.0.0`](/packages/numeric/arrow) [`audiocraft`](/packages/speech/audiocraft) [`awq:0.1.0`](/packages/llm/awq) [`bitsandbytes:0.39.1`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.4`](/packages/llm/bitsandbytes) [`bitsandbytes:0.45.5`](/packages/llm/bitsandbytes) [`bitsandbytes:0.46.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.47.0`](/packages/llm/bitsandbytes) [`bitsandbytes:0.48.0`](/packages/llm/bitsandbytes) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`cache_dit:0.2.0`](/packages/diffusion/cache_edit) [`canbus`](/packages/hw/canbus) [`cangaroo`](/packages/hw/cangaroo) [`causalconv1d:1.6.0`](/packages/ml/mamba/causalconv1d) [`chatterbox-tts`](/packages/speech/chatterbox-tts) [`ciso8601:v2.3.2`](/packages/smart-home/dependencies/ciso8601) [`clip_trt`](/packages/vit/clip_trt) [`cmake:pip`](/packages/build/cmake/cmake_pip) [`cobra:0.0.1`](/packages/ml/mamba/cobra) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`crossformer`](/packages/physicalAI/vla/crossformer) [`ctranslate2:4.6.0`](/packages/ml/ctranslate2) [`ctranslate2:master`](/packages/ml/ctranslate2) [`cuda-python:12.6`](/packages/cuda/cuda-python) [`cuda:12.6`](/packages/cuda/cuda) [`cuda:12.6-samples`](/packages/cuda/cuda) [`cudf:23.10.03`](/packages/ml/rapids/cudf) [`cudf:25.08.00`](/packages/ml/rapids/cudf) [`cudnn:9.3`](/packages/cuda/cudnn) [`cudnn:9.4`](/packages/cuda/cudnn) [`cuml`](/packages/ml/rapids/cuml) [`cupy`](/packages/numeric/cupy) [`cutlass:4.1.0`](/packages/cuda/cutlass) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`cv-cuda:0.15-cpp`](/packages/cv/cv-cuda) [`decord2:1.0.0`](/packages/multimedia/decord) [`deepspeed-kernels:0.1.0`](/packages/llm/deepspeed/deepspeed-kernels) [`deepspeed:0.15.2`](/packages/llm/deepspeed) [`deepspeed:0.18.0`](/packages/llm/deepspeed) [`deepspeed:0.9.5`](/packages/llm/deepspeed) [`deepstream`](/packages/cv/deepstream) [`diffusers:0.35.0`](/packages/diffusion/diffusers) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dimba:1.0`](/packages/ml/mamba/dimba) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`docker`](/packages/build/docker) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`easyvolcap:0.0.0`](/packages/3d/gaussian_splatting/easyvolcap) [`efficientvit`](/packages/vit/efficientvit) [`exllama:0.1`](/packages/llm/exllama) [`faiss:1.12.0`](/packages/vectordb/faiss) [`faiss_lite`](/packages/vectordb/faiss_lite) [`fast_gauss:1.0.0`](/packages/3d/gaussian_splatting/fast_gauss) [`faster-whisper`](/packages/speech/faster-whisper) [`ffmpeg:7.1`](/packages/multimedia/ffmpeg) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flashinfer:0.2.1.post2`](/packages/attention/flash-infer) [`flashinfer:0.2.2`](/packages/attention/flash-infer) [`flashinfer:0.2.2.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.6.post1`](/packages/attention/flash-infer) [`flashinfer:0.2.7`](/packages/attention/flash-infer) [`flashinfer:0.2.8`](/packages/attention/flash-infer) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`gdrcopy`](/packages/cuda/gdrcopy) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`genesis-world:0.2.2`](/packages/sim/genesis) [`glomap:2.0.0`](/packages/3d/3dvision/glomap) [`gptqmodel:3.0.1`](/packages/llm/gptqmodel) [`graphiti`](/packages/rag/graphiti) [`gsplat:1.5.3`](/packages/3d/gaussian_splatting/gsplat) [`gstreamer`](/packages/multimedia/gstreamer) [`h5py`](/packages/build/h5py) [`habitat-sim:0.3.4`](/packages/sim/habitat-sim) [`hloc:1.4`](/packages/3d/3dvision/hloc) [`hloc:1.5`](/packages/3d/3dvision/hloc) [`holoscan`](/packages/cv/holoscan) [`homeassistant-core:2025.7.0`](/packages/smart-home/homeassistant-core) [`huggingface_hub`](/packages/llm/huggingface_hub) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`isaac-ros:common-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:common-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:compression-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:data-tools-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:dnn-inference-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:image-pipeline-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:manipulator-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nitros-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:nvblox-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:pose-estimation-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-humble-desktop`](/packages/robots/isaac-ros) [`isaac-ros:visual-slam-3.2-jazzy-desktop`](/packages/robots/isaac-ros) [`isaaclab:2.2.0`](/packages/sim/isaac-sim/isaac-lab) [`isaacsim:5.0.0`](/packages/sim/isaac-sim) [`jax:0.4.38`](/packages/ml/jax) [`jax:0.6.3`](/packages/ml/jax) [`jetcam`](/packages/hw/jetcam) [`jetson-copilot`](/packages/rag/jetson-copilot) [`jetson-inference:foxy`](/packages/cv/jetson-inference) [`jetson-inference:galactic`](/packages/cv/jetson-inference) [`jetson-inference:humble`](/packages/cv/jetson-inference) [`jetson-inference:iron`](/packages/cv/jetson-inference) [`jetson-inference:jazzy`](/packages/cv/jetson-inference) [`jetson-inference:main`](/packages/cv/jetson-inference) [`jetson-utils:python`](/packages/multimedia/jetson-utils) [`jetson-utils:v1`](/packages/multimedia/jetson-utils) [`jetson-utils:v2`](/packages/multimedia/jetson-utils) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`kai_scheduler:0.5.5`](/packages/llm/dynamo/kai-scheduler) [`kaolin:0.18.0`](/packages/3d/3dvision/kaolin) [`kat:1`](/packages/ml/kans/kat) [`kokoro-tts:fastapi`](/packages/speech/kokoro-tts/kokoro-tts-fastapi) [`kokoro-tts:hf`](/packages/speech/kokoro-tts/kokoro-tts-hf) [`kokoro-tts:onnx`](/packages/speech/kokoro-tts/kokoro-tts-onnx) [`ktransformers:0.3.3`](/packages/llm/ktransformers) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain`](/packages/rag/langchain) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`libcom:0.1.0`](/packages/multimedia/libcom) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llama-index`](/packages/rag/llama-index) [`llama-index:samples`](/packages/rag/llama-index) [`llama-vision`](/packages/vlm/llama-vision) [`llama_cpp:0.2.57`](/packages/llm/llama_cpp) [`llama_cpp:0.2.70`](/packages/llm/llama_cpp) [`llama_cpp:0.2.83`](/packages/llm/llama_cpp) [`llama_cpp:0.2.90`](/packages/llm/llama_cpp) [`llama_cpp:0.3.1`](/packages/llm/llama_cpp) [`llama_cpp:0.3.2`](/packages/llm/llama_cpp) [`llama_cpp:0.3.5`](/packages/llm/llama_cpp) [`llama_cpp:0.3.6`](/packages/llm/llama_cpp) [`llama_cpp:0.3.7`](/packages/llm/llama_cpp) [`llama_cpp:0.3.8`](/packages/llm/llama_cpp) [`llama_cpp:0.3.9`](/packages/llm/llama_cpp) [`llama_cpp:0.4.0`](/packages/llm/llama_cpp) [`llama_cpp:b5255`](/packages/llm/llama_cpp) [`llama_cpp:b5833`](/packages/llm/llama_cpp) [`llamaspeak`](/packages/llm/llamaspeak) [`llava`](/packages/vlm/llava) [`llvm:10`](/packages/build/llvm) [`llvm:11`](/packages/build/llvm) [`llvm:12`](/packages/build/llvm) [`llvm:13`](/packages/build/llvm) [`llvm:14`](/packages/build/llvm) [`llvm:15`](/packages/build/llvm) [`llvm:16`](/packages/build/llvm) [`llvm:17`](/packages/build/llvm) [`llvm:18`](/packages/build/llvm) [`llvm:19`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`llvm:22`](/packages/build/llvm) [`lobechat`](/packages/llm/lobe_chat) [`local_llm`](/packages/llm/local_llm) [`log-linear-attention:0.0.1`](/packages/attention/log-linear-attention) [`mamba:2.2.5`](/packages/ml/mamba/mamba) [`mambavision:1.0`](/packages/ml/mamba/mambavision) [`memvid:0.1.4`](/packages/rag/memvid) [`meshlab:MeshLab-2023.12`](/packages/3d/3dvision/meshlab) [`meshlab:MeshLab-2025.03`](/packages/3d/3dvision/meshlab) [`mimicgen`](/packages/sim/mimicgen) [`minference:0.1.7`](/packages/llm/minference) [`minigpt4`](/packages/vlm/minigpt4) [`mistral_rs`](/packages/llm/mistral_rs) [`mlc:0.1.0`](/packages/llm/mlc) [`mlc:0.1.1`](/packages/llm/mlc) [`mlc:0.1.2`](/packages/llm/mlc) [`mlc:0.1.3`](/packages/llm/mlc) [`mlc:0.1.4`](/packages/llm/mlc) [`mlc:0.19.0`](/packages/llm/mlc) [`mlc:0.20.0`](/packages/llm/mlc) [`mlc:0.21.0`](/packages/llm/mlc) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`mooncake:0.3.5`](/packages/llm/dynamo/mooncake) [`mujoco`](/packages/sim/mujoco) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`nanodb`](/packages/vectordb/nanodb) [`nanoowl`](/packages/vit/nanoowl) [`nanosam`](/packages/vit/nanosam) [`nccl`](/packages/cuda/nccl) [`nemo`](/packages/llm/nemo) [`nerfacc:0.5.3`](/packages/3d/3dvision/nerfacc) [`nerfacc:0.5.4`](/packages/3d/3dvision/nerfacc) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`nerfview:0.1.4`](/packages/3d/gaussian_splatting/nerfview) [`ninja`](/packages/build/ninja) [`nixl:0.3.2`](/packages/llm/dynamo/nixl) [`numba`](/packages/numeric/numba) [`numpy`](/packages/numeric/numpy) [`nvdiffrast:0.3.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer/nvdiffrast) [`nvidia_modelopt:0.32.0`](/packages/llm/tensorrt_optimizer/nvidia-modelopt) [`nvpl`](/packages/cuda/nvpl) [`octo`](/packages/physicalAI/vla/octo) [`ollama:0.10.0`](/packages/llm/ollama) [`ollama:0.4.0`](/packages/llm/ollama) [`ollama:0.5.1`](/packages/llm/ollama) [`ollama:0.5.5`](/packages/llm/ollama) [`ollama:0.5.7`](/packages/llm/ollama) [`ollama:0.6.7`](/packages/llm/ollama) [`ollama:0.7.0`](/packages/llm/ollama) [`ollama:0.8.0`](/packages/llm/ollama) [`ollama:0.9.6`](/packages/llm/ollama) [`onnx`](/packages/ml/onnx) [`onnxruntime:1.19.2`](/packages/ml/onnxruntime) [`onnxruntime:1.20`](/packages/ml/onnxruntime) [`onnxruntime:1.20.1`](/packages/ml/onnxruntime) [`onnxruntime:1.21`](/packages/ml/onnxruntime) [`onnxruntime:1.22`](/packages/ml/onnxruntime) [`onnxruntime_genai:0.8.5`](/packages/ml/onnxruntime_genai) [`open3d:1.19.0`](/packages/3d/3dvision/open3d) [`openai`](/packages/llm/openai) [`opencv:4.10.0`](/packages/cv/opencv) [`opencv:4.10.0-meta`](/packages/cv/opencv) [`opencv:4.11.0`](/packages/cv/opencv) [`opencv:4.11.0-meta`](/packages/cv/opencv) [`opencv:4.12.0`](/packages/cv/opencv) [`opencv:4.12.0-meta`](/packages/cv/opencv) [`opencv:4.8.1`](/packages/cv/opencv) [`opencv:4.8.1-deb`](/packages/cv/opencv) [`opencv:4.8.1-meta`](/packages/cv/opencv) [`opendronemap`](/packages/robots/opendronemap) [`opendronemap:node`](/packages/robots/opendronemap) [`openpi`](/packages/robots/openpi) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`optimum`](/packages/llm/optimum) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`physx`](/packages/sim/physx) [`piper-tts`](/packages/speech/piper-tts) [`piper1-tts:1.3.0`](/packages/speech/piper1-tts) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`polyscope:2.5.0`](/packages/3d/3dvision/polyscope) [`prismatic`](/packages/vlm/prismatic) [`protobuf:apt`](/packages/build/protobuf/protobuf_apt) [`protobuf:cpp`](/packages/build/protobuf/protobuf_cpp) [`protomotions:2.5.0`](/packages/robots/protomotions) [`psutil-home-assistant:0.0.1`](/packages/smart-home/dependencies/psutil-home-assistant) [`pyav`](/packages/multimedia/pyav) [`pybind11`](/packages/build/pybind11) [`pyceres:2.5`](/packages/3d/3dvision/pyceres) [`pycolmap:3.12`](/packages/3d/3dvision/pycolmap) [`pycolmap:3.13`](/packages/3d/3dvision/pycolmap) [`pycuda`](/packages/cuda/pycuda) [`pykan:0.2.9`](/packages/ml/kans/pykan) [`pymeshlab:2023.12.post2`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2023.12.post3`](/packages/3d/3dvision/pymeshlab) [`pymeshlab:2025.6.23.dev0`](/packages/3d/3dvision/pymeshlab) [`python:3.10`](/packages/build/python) [`python:3.11`](/packages/build/python) [`python:3.12`](/packages/build/python) [`python:3.13`](/packages/build/python) [`python:3.14`](/packages/build/python) [`pytorch3d:0.7.8`](/packages/pytorch/torch3d) [`pytorch3d:0.7.9`](/packages/pytorch/torch3d) [`pytorch:2.1`](/packages/pytorch) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`radial-attention:0.1.0`](/packages/attention/radial-attention) [`raft`](/packages/ml/rapids/raft) [`realsense`](/packages/hw/realsense) [`riva-client:python`](/packages/speech/riva-client) [`robogen`](/packages/sim/robogen) [`robomimic`](/packages/sim/robomimic) [`robopoint`](/packages/physicalAI/vla/robopoint) [`robosuite`](/packages/sim/robosuite) [`ros:foxy-desktop`](/packages/robots/ros) [`ros:foxy-ros-base`](/packages/robots/ros) [`ros:foxy-ros-core`](/packages/robots/ros) [`ros:galactic-desktop`](/packages/robots/ros) [`ros:galactic-ros-base`](/packages/robots/ros) [`ros:galactic-ros-core`](/packages/robots/ros) [`ros:humble-desktop`](/packages/robots/ros) [`ros:humble-ros-base`](/packages/robots/ros) [`ros:humble-ros-core`](/packages/robots/ros) [`ros:iron-desktop`](/packages/robots/ros) [`ros:iron-ros-base`](/packages/robots/ros) [`ros:iron-ros-core`](/packages/robots/ros) [`ros:jazzy-desktop`](/packages/robots/ros) [`ros:jazzy-ros-base`](/packages/robots/ros) [`ros:jazzy-ros-core`](/packages/robots/ros) [`ros:noetic-desktop`](/packages/robots/ros) [`ros:noetic-ros-base`](/packages/robots/ros) [`ros:noetic-ros-core`](/packages/robots/ros) [`rust`](/packages/build/rust) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sam`](/packages/vit/sam) [`sapien`](/packages/sim/sapien) [`sapiens`](/packages/vit/sapiens) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sound-utils`](/packages/multimedia/sound-utils) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`spark-tts`](/packages/speech/spark-tts) [`speaches`](/packages/speech/speaches) [`splashsurf`](/packages/sim/genesis/splashSurf) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`sudonim`](/packages/llm/sudonim) [`sudonim:hf`](/packages/llm/sudonim) [`quadrants:1.8.0`](/packages/sim/genesis/quadrants) [`tam`](/packages/vit/tam) [`tensorboard`](/packages/ml/tensorflow/tensorboard) [`tensorflow2:2.16.1`](/packages/ml/tensorflow) [`tensorflow2:2.18.0`](/packages/ml/tensorflow) [`tensorflow2:2.19.0`](/packages/ml/tensorflow) [`tensorflow2:2.20.0`](/packages/ml/tensorflow) [`tensorflow2:2.21.0`](/packages/ml/tensorflow) [`tensorflow_graphics:2.18.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.19.0`](/packages/ml/tensorflow/graphics) [`tensorflow_graphics:2.20.0`](/packages/ml/tensorflow/graphics) [`tensorflow_text:2.18.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.19.0`](/packages/ml/tensorflow/text) [`tensorflow_text:2.20.0`](/packages/ml/tensorflow/text) [`tensorrt:10.3`](/packages/cuda/tensorrt) [`tensorrt:10.4`](/packages/cuda/tensorrt) [`tensorrt:10.5`](/packages/cuda/tensorrt) [`tensorrt:10.7`](/packages/cuda/tensorrt) [`tensorrt_llm:0.12`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`tensorrt_llm:0.22.0`](/packages/llm/tensorrt_optimizer/tensorrt_llm) [`text-generation-inference`](/packages/llm/text-generation-inference) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`tinycudann:1.7`](/packages/3d/3dvision/tinycudann) [`torch-memory-saver:0.0.7`](/packages/pytorch/torchsaver) [`torch2trt`](/packages/pytorch/torch2trt) [`torch_tensorrt`](/packages/pytorch/torch_tensorrt) [`torchao:0.12.0`](/packages/pytorch/torchao) [`torchao:0.13.0`](/packages/pytorch/torchao) [`torchaudio:2.1.0`](/packages/pytorch/torchaudio) [`torchaudio:2.2.2`](/packages/pytorch/torchaudio) [`torchaudio:2.3.0`](/packages/pytorch/torchaudio) [`torchaudio:2.4.0`](/packages/pytorch/torchaudio) [`torchaudio:2.5.0`](/packages/pytorch/torchaudio) [`torchaudio:2.6.0`](/packages/pytorch/torchaudio) [`torchaudio:2.7.0`](/packages/pytorch/torchaudio) [`torchaudio:2.8.0`](/packages/pytorch/torchaudio) [`torchsde:0.2.1`](/packages/pytorch/torchsde) [`torchsde:0.2.2`](/packages/pytorch/torchsde) [`torchsde:0.2.3`](/packages/pytorch/torchsde) [`torchsde:0.2.4`](/packages/pytorch/torchsde) [`torchsde:0.2.5`](/packages/pytorch/torchsde) [`torchsde:0.2.6`](/packages/pytorch/torchsde) [`torchsde:0.2.7`](/packages/pytorch/torchsde) [`torchsde:0.2.8`](/packages/pytorch/torchsde) [`torchtext:0.18.0`](/packages/pytorch/torchtext) [`torchtext:0.19.0`](/packages/pytorch/torchtext) [`torchvision:0.16.2`](/packages/pytorch/torchvision) [`torchvision:0.17.2`](/packages/pytorch/torchvision) [`torchvision:0.18.0`](/packages/pytorch/torchvision) [`torchvision:0.19.1`](/packages/pytorch/torchvision) [`torchvision:0.20.0`](/packages/pytorch/torchvision) [`torchvision:0.21.0`](/packages/pytorch/torchvision) [`torchvision:0.22.0`](/packages/pytorch/torchvision) [`torchvision:0.23.0`](/packages/pytorch/torchvision) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`transformers:4.53.1`](/packages/llm/transformers) [`triton:3.0.0`](/packages/ml/triton) [`triton:3.1.0`](/packages/ml/triton) [`triton:3.2.0`](/packages/ml/triton) [`triton:3.3.0`](/packages/ml/triton) [`triton:3.3.1`](/packages/ml/triton) [`triton:3.4.0`](/packages/ml/triton) [`triton:3.5.0`](/packages/ml/triton) [`tritonserver`](/packages/ml/tritonserver) [`tvm`](/packages/ml/tvm) [`video-codec-sdk:12.2.72`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:12.2.72-samples`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19`](/packages/multimedia/video-codec-sdk) [`video-codec-sdk:13.0.19-samples`](/packages/multimedia/video-codec-sdk) [`videollama:1.0.0`](/packages/vlm/videollama) [`videomambasuite:1.0`](/packages/ml/mamba/videomambasuite) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`vpi`](/packages/cv/vpi) [`vscode`](/packages/code/vscode) [`vscode:cuda`](/packages/code/vscode) [`vscode:torch`](/packages/code/vscode) [`vscode:transformers`](/packages/code/vscode) [`vtk:9.3.1`](/packages/sim/genesis/vtk) [`vtk:9.4.2`](/packages/sim/genesis/vtk) [`vtk:9.5.0`](/packages/sim/genesis/vtk) [`vulkan:1.3.275`](/packages/multimedia/vulkan) [`vulkan:1.3.296`](/packages/multimedia/vulkan) [`vulkan:1.4.309`](/packages/multimedia/vulkan) [`vulkan:1.4.313`](/packages/multimedia/vulkan) [`vulkan:1.4.314`](/packages/multimedia/vulkan) [`warp:1.7.0`](/packages/numeric/warp) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.7.0-jax`](/packages/numeric/warp) [`warp:1.7.0-torch`](/packages/numeric/warp) [`warp:1.8.1`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`warp:1.8.1-jax`](/packages/numeric/warp) [`warp:1.8.1-torch`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) [`whisper_trt`](/packages/speech/whisper_trt) [`whisperx`](/pa\n\n<!-- [TRUNCATED AT 50KB] -->\n\n<!-- File: packages/cuda/cudastack/README.md (Tokens: ~1829 | Category: plugin-manifest) -->\n# CUDA Stack - Consolidated Library Installation\n\n## What is cudastack?\n\n`cudastack` is a **consolidation package** that installs multiple CUDA libraries (cuDNN, NCCL, TensorRT, CUTLASS, etc.) in **ONE RUN command** to avoid Docker's \"max depth exceeded\" error.\n\n## Important: cudastack vs cuda\n\n```\ncuda/ cudastack/\n├── Dockerfile ├── Dockerfile (ONE RUN for all libs)\n├── Dockerfile.pip ├── config.py (variants)\n├── Dockerfile.builtin └── install/ & build/ scripts\n├── Dockerfile.samples\n└── config.py\n ├── cuda_package() cudastack DEPENDS ON cuda\n ├── cuda_builtin() (doesn't replace it!)\n ├── cuda_samples()\n └── pip_cache()\n```\n\n**Key Points:**\n\n1. ✅ **`cuda` package** - Installs CUDA toolkit, provides pip_cache, samples, builtin variants\n2. ✅ **`cudastack` package** - Builds ON TOP of `cuda`, adds cuDNN/NCCL/TensorRT/etc. in ONE layer\n3. ✅ **pip_cache, cuda_samples, cuda_builtin** - Still work! They're part of `cuda`, not `cudastack`\n\n## The Problem It Solves\n\n### Before (Separate Packages):\n```\nLayer 1: cuda\nLayer 2: cudnn\nLayer 3: tensorrt\nLayer 4: nccl\nLayer 5: cutlass\n...\nLayer 50+: \"max depth exceeded\" ❌\n```\n\n### After (cudastack):\n```\nLayer 1: cuda (with pip_cache, samples, etc.)\nLayer 2: cudastack (cudnn + nccl + tensorrt + cutlass in ONE RUN)\nLayer 3: your-app\nResult: Works! ✅\n```\n\n## Architecture\n\n```\nBASE_IMAGE\n ↓\ncuda:12.4 (from cuda/Dockerfile)\n ├── pip_cache:cu124 (from cuda/Dockerfile.pip)\n └── cuda:12.4-samples (from cuda/Dockerfile.samples)\n ↓\ncudastack:minimal (from cudastack/Dockerfile)\n ├── cuDNN\n ├── NCCL\n └── (all installed in ONE RUN)\n ↓\nyour-app\n```\n\n## Variants\n\n| Variant | Includes | Use Case |\n|---------|----------|----------|\n| `cudastack:minimal` | cuDNN + NCCL | Training, inference |\n| `cudastack:standard` | + TensorRT | Optimized inference |\n| `cudastack:full` | + CUTLASS + GDRCopy | Research, development |\n\n## Usage\n\n### Build with jetson-containers:\n\n```bash\n# Build minimal stack\n./build.sh cudastack:minimal\n\n# Build standard (with TensorRT)\n./build.sh cudastack:standard\n\n# Build full (with CUTLASS)\n./build.sh cudastack:full\n\n# Use as base for your app\n./build.sh --base cudastack:minimal pytorch\n```\n\n### The cudastack Dockerfile does ONE RUN:\n\n```dockerfile\n# Copy all install/build scripts\nCOPY install/*.sh /tmp/cudastack/install/\nCOPY build/*.sh /tmp/cudastack/build/\n\n# ONE RUN command calls all scripts conditionally\nRUN set -ex && \\\n if [ \"$WITH_CUDNN\" = \"1\" ]; then \\\n ./install/install_cudnn.sh; \\\n fi && \\\n if [ \"$WITH_NCCL\" = \"1\" ]; then \\\n ./install/install_nccl.sh; \\\n fi && \\\n if [ \"$WITH_TENSORRT\" = \"1\" ]; then \\\n ./install/install_tensorrt.sh; \\\n fi && \\\n # ... more components ...\n rm -rf /tmp/cudastack\n```\n\n**Key**: All library installations happen in this ONE RUN = ONE Docker layer!\n\n## Setup\n\nThe install scripts need to be populated from existing packages:\n\n```bash\ncd packages/cuda/cudastack\n./setup-scripts.sh\n```\n\nThen review and complete the scripts in `install/` and `build/`.\n\n## What About pip_cache, samples, builtin?\n\nThese are **still available** from the `cuda` package:\n\n```bash\n# pip_cache is automatically included by cuda package\n./build.sh cuda:12.4\n# This creates both:\n# - cuda:12.4 (main CUDA toolkit)\n# - pip_cache:cu124 (pip caching)\n\n# CUDA samples work as before\n./build.sh cuda:12.4-samples\n\n# Builtin works as before (for JetPack 4-5)\n./build.sh cuda:11.4 # Uses Dockerfile.builtin when CUDA pre-installed\n```\n\nThe `cuda` package's `config.py` returns multiple packages:\n```python\n# In cuda/config.py\ndef cuda_package(version, url, ...):\n cuda = package.copy()\n cuda['name'] = f'cuda:{version}'\n \n cuda_pip = pip_cache(version, requires) # Creates pip_cache package\n \n return cuda, cuda_pip # Returns TWO packages!\n```\n\n**cudastack doesn't need to duplicate this** - it just depends on `cuda` being installed first.\n\n## Component Toggles\n\nControlled by `WITH_*` build args (set in config.py):\n\n- ✅ `WITH_CUDNN=1` - cuDNN (always on)\n- ✅ `WITH_NCCL=1` - NCCL (always on)\n- 🎛️ `WITH_TENSORRT` - TensorRT (on in standard/full)\n- 🎛️ `WITH_CUTLASS` - CUTLASS (on in full only)\n- 🎛️ `WITH_GDRCOPY` - GDRCopy (on for Tegra)\n- ⚙️ `WITH_CUDSS=0` - cuDSS (off by default)\n- ⚙️ `WITH_CUSPARSELT=0` - cuSPARSELt (off by default)\n- ⚙️ `WITH_CUTENSOR=0` - cuTENSOR (off by default)\n- ⚙️ `WITH_NVPL=0` - NVPL (off by default)\n- ⚙️ `WITH_NVSHMEM=0` - NVSHMEM (off by default)\n\n## How config.py Works\n\nFollowing the jetson-containers pattern:\n\n```python\ndef cuda_stack(name, with_tensorrt=False, with_cutlass=False, requires=None):\n \"\"\"Generate a cudastack package variant\"\"\"\n pkg = package.copy()\n pkg['name'] = name\n pkg['depends'] = ['cuda'] # Depends ON cuda, doesn't replace it\n \n pkg['build_args'] = {\n 'WITH_CUDNN': '1',\n 'WITH_TENSORRT': '1' if with_tensorrt else '0',\n 'WITH_NCCL': '1',\n # ... etc\n }\n \n return pkg\n\n# Define variants\npackage = [\n cuda_stack('cudastack:minimal', with_tensorrt=False),\n cuda_stack('cudastack:standard', with_tensorrt=True),\n cuda_stack('cudastack:full', with_tensorrt=True, with_cutlass=True),\n]\n```\n\n## Comparison with Individual Packages\n\n| Approach | Layers | Build Time | Complexity |\n|----------|--------|------------|------------|\n| **Individual packages** | 50+ | Longer | Simple per-package |\n| **cudastack (consolidated)** | 2-3 | Faster | One complex layer |\n\nBoth approaches work! Use `cudastack` when you need multiple libraries and want to avoid layer limits.\n\n## See Also\n\n- [cuda package](../cuda/) - CUDA toolkit, pip_cache, samples, builtin\n- [cudnn package](../cudnn/) - Individual cuDNN package\n- [tensorrt package](../tensorrt/) - Individual TensorRT package (has Dockerfile.deb and Dockerfile.tar)\n- [nccl package](../nccl/) - Individual NCCL package\n\n## Summary\n\n✅ **cudastack is a consolidation layer** on top of cuda \n✅ **pip_cache, samples, builtin still work** from the cuda package \n✅ **ONE RUN command** installs all libraries to avoid layer limits \n✅ **Follows jetson-containers patterns** with config.py and variants\n\n<!-- File: packages/cuda/cutlass/README.md (Tokens: ~2912 | Category: plugin-manifest) -->\n# cutlass\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`cutlass:4.1.0`** | |\n| :-- | :-- |\n|    Aliases | `cutlass` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cuda-python`](/packages/cuda/cuda-python) [`cudnn`](/packages/cuda/cudnn) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch`](/packages/pytorch) |\n|    Dependants | [`awq:0.1.0`](/packages/llm/awq) [`block_sparse_attn:0.0.1`](/packages/attention/block-sparse-attention) [`comfyui`](/packages/diffusion/comfyui) [`cosmos-predict2`](/packages/diffusion/cosmos/cosmos-predict2) [`cosmos-reason1`](/packages/diffusion/cosmos/cosmos-reason1) [`cosmos-transfer1`](/packages/diffusion/cosmos/cosmos-transfer1) [`cosmos1-diffusion-renderer:1.0.4`](/packages/diffusion/cosmos/cosmos_diffusion_renderer) [`dynamo:0.3.2`](/packages/llm/dynamo/dynamo) [`exllama:0.1`](/packages/llm/exllama) [`flash-attention:2.5.7`](/packages/attention/flash-attention) [`flash-attention:2.6.3`](/packages/attention/flash-attention) [`flash-attention:2.7.2.post1`](/packages/attention/flash-attention) [`flash-attention:2.7.4.post1`](/packages/attention/flash-attention) [`flash-attention:2.8.0.post2`](/packages/attention/flash-attention) [`flash-attention:2.8.1`](/packages/attention/flash-attention) [`flexprefill:0.1.0`](/packages/attention/flexprefill) [`framepack`](/packages/diffusion/framepack) [`genai-bench:0.1.0`](/packages/llm/sglang/genai-bench) [`huggingface_kernels:0.7.0`](/packages/attention/huggingface_kernels) [`hymba`](/packages/llm/hymba) [`isaac-gr00t`](/packages/physicalAI/vla/isaac-gr00t) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`lita`](/packages/vlm/lita) [`llama-factory`](/packages/llm/llama-factory) [`llava`](/packages/vlm/llava) [`lobechat`](/packages/llm/lobe_chat) [`minference:0.1.7`](/packages/llm/minference) [`mlstm_kernels:2.0.1`](/packages/ml/xlstm/mlstm_kernels) [`nano_llm:24.4`](/packages/llm/nano_llm) [`nano_llm:24.4-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4-humble`](/packages/llm/nano_llm) [`nano_llm:24.4-iron`](/packages/llm/nano_llm) [`nano_llm:24.4.1`](/packages/llm/nano_llm) [`nano_llm:24.4.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.4.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.4.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.4.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.5`](/packages/llm/nano_llm) [`nano_llm:24.5-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5-humble`](/packages/llm/nano_llm) [`nano_llm:24.5-iron`](/packages/llm/nano_llm) [`nano_llm:24.5.1`](/packages/llm/nano_llm) [`nano_llm:24.5.1-foxy`](/packages/llm/nano_llm) [`nano_llm:24.5.1-galactic`](/packages/llm/nano_llm) [`nano_llm:24.5.1-humble`](/packages/llm/nano_llm) [`nano_llm:24.5.1-iron`](/packages/llm/nano_llm) [`nano_llm:24.6`](/packages/llm/nano_llm) [`nano_llm:24.6-foxy`](/packages/llm/nano_llm) [`nano_llm:24.6-galactic`](/packages/llm/nano_llm) [`nano_llm:24.6-humble`](/packages/llm/nano_llm) [`nano_llm:24.6-iron`](/packages/llm/nano_llm) [`nano_llm:24.7`](/packages/llm/nano_llm) [`nano_llm:24.7-foxy`](/packages/llm/nano_llm) [`nano_llm:24.7-galactic`](/packages/llm/nano_llm) [`nano_llm:24.7-humble`](/packages/llm/nano_llm) [`nano_llm:24.7-iron`](/packages/llm/nano_llm) [`nano_llm:main`](/packages/llm/nano_llm) [`nano_llm:main-foxy`](/packages/llm/nano_llm) [`nano_llm:main-galactic`](/packages/llm/nano_llm) [`nano_llm:main-humble`](/packages/llm/nano_llm) [`nano_llm:main-iron`](/packages/llm/nano_llm) [`openvla`](/packages/physicalAI/vla/openvla) [`openvla:mimicgen`](/packages/physicalAI/vla/openvla) [`paraattention:0.4.0`](/packages/attention/ParaAttention) [`partpacker:0.1.0`](/packages/3d/3dobjects/partpacker) [`plstm:0.1.0`](/packages/ml/xlstm/pltsm) [`prismatic`](/packages/vlm/prismatic) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`robopoint`](/packages/physicalAI/vla/robopoint) [`sage-attention:3.0.0`](/packages/attention/sage-attention) [`sdnext`](/packages/diffusion/sdnext) [`self-forcing`](/packages/diffusion/self-forcing) [`sgl-kernel:0.2.3`](/packages/llm/sglang/sgl-kernel) [`sglang:0.4.4`](/packages/llm/sglang) [`sglang:0.4.6`](/packages/llm/sglang) [`sglang:0.4.9`](/packages/llm/sglang) [`sparc3d:0.1.0`](/packages/3d/3dobjects/sparc3d) [`sparge-attention:0.1.0`](/packages/attention/sparge-attention) [`sudonim:hf`](/packages/llm/sudonim) [`text-generation-webui:1.7`](/packages/llm/text-generation-webui) [`text-generation-webui:6a7cd01`](/packages/llm/text-generation-webui) [`text-generation-webui:main`](/packages/llm/text-generation-webui) [`transformer-engine:2.7`](/packages/ml/transformer-engine) [`videollama:1.0.0`](/packages/vlm/videollama) [`vila`](/packages/vlm/vila) [`vllm:0.7.4`](/packages/llm/vllm) [`vllm:0.8.4`](/packages/llm/vllm) [`vllm:0.9.0`](/packages/llm/vllm) [`vllm:0.9.2`](/packages/llm/vllm) [`vllm:0.9.3`](/packages/llm/vllm) [`vllm:v0.8.5.post1`](/packages/llm/vllm) [`vscode:transformers`](/packages/code/vscode) [`xattention:0.0.1`](/packages/attention/xattention) [`xlstm:2.0.5`](/packages/ml/xlstm/xlstm) [`xtuner`](/packages/vlm/xtuner) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag cutlass)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host cutlass:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag cutlass)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag cutlass) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build cutlass\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cuda/pycuda/README.md (Tokens: ~2214 | Category: plugin-manifest) -->\n# pycuda\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`pycuda`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=32.6']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda`](/packages/cuda/cuda) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) |\n|    Dependants | [`audiocraft`](/packages/speech/audiocraft) [`cv-cuda:0.15`](/packages/cv/cv-cuda) [`diffusion_policy`](/packages/diffusion/diffusion_policy) [`dli-nano-ai`](/packages/ml/dli/dli-nano-ai) [`efficientvit`](/packages/vit/efficientvit) [`jupyter_clickable_image_widget`](/packages/hw/jupyter_clickable_image_widget) [`jupyterlab:4.2.0`](/packages/code/jupyterlab) [`jupyterlab:4.2.0-myst`](/packages/code/jupyterlab) [`jupyterlab:latest`](/packages/code/jupyterlab) [`jupyterlab:latest-myst`](/packages/code/jupyterlab) [`l4t-diffusion`](/packages/ml/l4t/l4t-diffusion) [`l4t-dynamo`](/packages/ml/l4t/l4t-dynamo) [`l4t-ml`](/packages/ml/l4t/l4t-ml) [`l4t-pytorch`](/packages/ml/l4t/l4t-pytorch) [`l4t-tensorflow:tf1`](/packages/ml/l4t/l4t-tensorflow) [`l4t-tensorflow:tf2`](/packages/ml/l4t/l4t-tensorflow) [`l4t-text-generation`](/packages/ml/l4t/l4t-text-generation) [`langchain:samples`](/packages/rag/langchain) [`lerobot`](/packages/robots/lerobot) [`llama-index:samples`](/packages/rag/llama-index) [`openpi`](/packages/robots/openpi) [`pytorch:2.1-all`](/packages/pytorch) [`pytorch:2.2-all`](/packages/pytorch) [`pytorch:2.3-all`](/packages/pytorch) [`pytorch:2.3.1-all`](/packages/pytorch) [`pytorch:2.4-all`](/packages/pytorch) [`pytorch:2.5-all`](/packages/pytorch) [`pytorch:2.6-all`](/packages/pytorch) [`pytorch:2.7-all`](/packages/pytorch) [`pytorch:2.8-all`](/packages/pytorch) [`sam`](/packages/vit/sam) [`stable-diffusion-webui`](/packages/diffusion/stable-diffusion-webui) [`tam`](/packages/vit/tam) [`voice-pro`](/packages/speech/voice-pro) [`voicecraft`](/packages/speech/voicecraft) [`warp:1.7.0-all`](/packages/numeric/warp) [`warp:1.8.1-all`](/packages/numeric/warp) [`whisper`](/packages/speech/whisper) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Images | [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 0.4GB)`<br>[`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-09-07, 5.0GB)`<br>[`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-08-29, 5.0GB)`<br>[`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 5.0GB)`<br>[`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2023-12-06, 3.5GB)`<br>[`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) `(2025-03-03, 2.3GB)` |\n\n</details>\n\n<details open>\n<summary><b><a id=\"images\">CONTAINER IMAGES</a></b></summary>\n<br>\n\n| Repository/Tag | Date | Arch | Size |\n| :-- | :--: | :--: | :--: |\n|   [`dustynv/pycuda:r32.7.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `0.4GB` |\n|   [`dustynv/pycuda:r35.2.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-09-07` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.3.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-08-29` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r35.4.1`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `5.0GB` |\n|   [`dustynv/pycuda:r36.2.0`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2023-12-06` | `arm64` | `3.5GB` |\n|   [`dustynv/pycuda:r36.4.0-cu128-24.04`](https://hub.docker.com/r/dustynv/pycuda/tags) | `2025-03-03` | `arm64` | `2.3GB` |\n\n> <sub>Container images are compatible with other minor versions of JetPack/L4T:</sub><br>\n> <sub>    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)</sub><br>\n> <sub>    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)</sub><br>\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag pycuda)\n\n# or explicitly specify one of the container images above\njetson-containers run dustynv/pycuda:r36.4.0-cu128-24.04\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host dustynv/pycuda:r36.4.0-cu128-24.04\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag pycuda)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag pycuda) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build pycuda\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dobjects/partpacker/README.md (Tokens: ~1124 | Category: plugin-manifest) -->\n# partpacker\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`partpacker:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `partpacker` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVlabs/PartPacker |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag partpacker)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host partpacker:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag partpacker)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag partpacker) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build partpacker\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dobjects/sparc3d/README.md (Tokens: ~1136 | Category: plugin-manifest) -->\n# sparc3d\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`sparc3d:0.1.0`** | |\n| :-- | :-- |\n|    Aliases | `sparc3d` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn:9.3`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`triton`](/packages/ml/triton) [`tensorrt`](/packages/cuda/tensorrt) [`onnxruntime`](/packages/ml/onnxruntime) [`opengl`](/packages/multimedia/opengl) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`ffmpeg`](/packages/multimedia/ffmpeg) [`opencv`](/packages/cv/opencv) [`huggingface_hub`](/packages/llm/huggingface_hub) [`rust`](/packages/build/rust) [`transformers`](/packages/llm/transformers) [`diffusers`](/packages/diffusion/diffusers) [`xformers`](/packages/attention/xformers) [`cuda-python`](/packages/cuda/cuda-python) [`cutlass`](/packages/cuda/cutlass) [`flash-attention`](/packages/attention/flash-attention) [`ninja`](/packages/build/ninja) [`polyscope`](/packages/3d/3dvision/polyscope) [`pymeshlab`](/packages/3d/3dvision/pymeshlab) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/lizhihao6/Sparc3D |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag sparc3d)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host sparc3d:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag sparc3d)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag sparc3d) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build sparc3d\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dvision/glomap/README.md (Tokens: ~993 | Category: plugin-manifest) -->\n# glomap\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`glomap:2.0.0`** | |\n| :-- | :-- |\n|    Aliases | `glomap` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/colmap/colmap.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag glomap)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host glomap:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag glomap)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag glomap) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build glomap\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dvision/hloc/README.md (Tokens: ~1404 | Category: plugin-manifest) -->\n# hloc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`hloc:1.4`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n| **`hloc:1.5`** | |\n| :-- | :-- |\n|    Aliases | `hloc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`torch`](/packages/pytorch) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`ffmpeg`](/packages/multimedia/ffmpeg) [`ninja`](/packages/build/ninja) [`opengl`](/packages/multimedia/opengl) [`pyceres`](/packages/3d/3dvision/pyceres) [`pycolmap`](/packages/3d/3dvision/pycolmap) [`llvm`](/packages/build/llvm) [`vulkan`](/packages/multimedia/vulkan) [`video-codec-sdk`](/packages/multimedia/video-codec-sdk) [`opencv`](/packages/cv/opencv) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) [`pixsfm:1.0`](/packages/3d/3dvision/pixsfm) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cvg/Hierarchical-Localization.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag hloc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host hloc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag hloc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag hloc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build hloc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dvision/kaolin/README.md (Tokens: ~965 | Category: plugin-manifest) -->\n# kaolin\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`kaolin:0.18.0`** | |\n| :-- | :-- |\n|    Aliases | `kaolin` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) [`tinycudann`](/packages/3d/3dvision/tinycudann) [`opengl`](/packages/multimedia/opengl) [`polyscope`](/packages/3d/3dvision/polyscope) |\n|    Dependants | [`3dgrut:2.0.0`](/packages/3d/gaussian_splatting/3dgrut) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/NVIDIAGameWorks/kaolin |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag kaolin)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host kaolin:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag kaolin)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag kaolin) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build kaolin\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dvision/meshlab/README.md (Tokens: ~981 | Category: plugin-manifest) -->\n# meshlab\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`meshlab:MeshLab-2023.12`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n| **`meshlab:MeshLab-2025.03`** | |\n| :-- | :-- |\n|    Aliases | `meshlab` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache`](/packages/cuda/cuda) [`python`](/packages/build/python) [`cmake`](/packages/build/cmake/cmake_pip) [`ninja`](/packages/build/ninja) [`numpy`](/packages/numeric/numpy) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/cnr-isti-vclab/MeshLab.git |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag meshlab)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host meshlab:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag meshlab)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag meshlab) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build meshlab\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- File: packages/cv/3d/3dvision/nerfacc/README.md (Tokens: ~1178 | Category: plugin-manifest) -->\n# nerfacc\n\n> [`CONTAINERS`](#user-content-containers) [`IMAGES`](#user-content-images) [`RUN`](#user-content-run) [`BUILD`](#user-content-build)\n\n<details open>\n<summary><b><a id=\"containers\">CONTAINERS</a></b></summary>\n<br>\n\n| **`nerfacc:0.5.3`** | |\n| :-- | :-- |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n| **`nerfacc:0.5.4`** | |\n| :-- | :-- |\n|    Aliases | `nerfacc` |\n|    Requires | `L4T ['>=34.1.0']` |\n|    Dependencies | [`build-essential`](/packages/build/build-essential) [`pip_cache:cu126`](/packages/cuda/cuda) [`cuda:12.6`](/packages/cuda/cuda) [`cudnn`](/packages/cuda/cudnn) [`python`](/packages/build/python) [`numpy`](/packages/numeric/numpy) [`cmake`](/packages/build/cmake/cmake_pip) [`onnx`](/packages/ml/onnx) [`pytorch:2.8`](/packages/pytorch) [`torchvision`](/packages/pytorch/torchvision) [`torchaudio`](/packages/pytorch/torchaudio) |\n|    Dependants | [`fruitnerf:1.0`](/packages/3d/nerf/fruitnerf) [`nerfstudio:1.1.7`](/packages/3d/nerf/nerfstudio) |\n|    Dockerfile | [`Dockerfile`](Dockerfile) |\n|    Notes | https://github.com/nerfstudio-project/nerfacc |\n\n</details>\n\n<details open>\n<summary><b><a id=\"run\">RUN CONTAINER</a></b></summary>\n<br>\n\nTo start the container, you can use [`jetson-containers run`](/docs/run.md) and [`autotag`](/docs/run.md#autotag), or manually put together a [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) command:\n```bash\n# automatically pull or build a compatible container image\njetson-containers run $(autotag nerfacc)\n\n# or if using 'docker run' (specify image and mounts/ect)\nsudo docker run --runtime nvidia -it --rm --network=host nerfacc:36.4.0\n\n```\n> <sup>[`jetson-containers run`](/docs/run.md) forwards arguments to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/) with some defaults added (like `--runtime nvidia`, mounts a `/data` cache, and detects devices)</sup><br>\n> <sup>[`autotag`](/docs/run.md#autotag) finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.</sup>\n\nTo mount your own directories into the container, use the [`-v`](https://docs.docker.com/engine/reference/commandline/run/#volume) or [`--volume`](https://docs.docker.com/engine/reference/commandline/run/#volume) flags:\n```bash\njetson-containers run -v /path/on/host:/path/in/container $(autotag nerfacc)\n```\nTo launch the container running a command, as opposed to an interactive shell:\n```bash\njetson-containers run $(autotag nerfacc) my_app --abc xyz\n```\nYou can pass any options to it that you would to [`docker run`](https://docs.docker.com/engine/reference/commandline/run/), and it'll print out the full command that it constructs before executing it.\n</details>\n<details open>\n<summary><b><a id=\"build\">BUILD CONTAINER</b></summary>\n<br>\n\nIf you use [`autotag`](/docs/run.md#autotag) as shown above, it'll ask to build the container for you if needed. To manually build it, first do the [system setup](/docs/setup.md), then run:\n```bash\njetson-containers build nerfacc\n```\nThe dependencies from above will be built into the container, and it'll be tested during. Run it with [`--help`](/jetson_containers/build.py) for build options.\n</details>\n\n<!-- ... +227 more modular skills available via /api/skills/dusty-nv/jetson-containers?path={path} -->\n</yakaai_skills>"}