onnx

onnx

CONTAINERS IMAGES RUN BUILD

CONTAINERS

onnx
Builds
Requires	L4T >=32.6
Dependencies	build-essential python cmake numpy
Dependants	audiocraft auto_gptq awq awq:dev bitsandbytes efficientvit exllama:v1 exllama:v2 faiss:lite gptq-for-llama jetson-inference l4t-diffusion l4t-ml l4t-pytorch l4t-text-generation langchain langchain:samples llava local_llm minigpt4 mlc:51fb0f4 mlc:9bf5723 mlc:dev nanodb nanoowl nanosam nemo onnxruntime optimum pytorch:1.10 pytorch:1.11 pytorch:1.12 pytorch:1.13 pytorch:1.9 pytorch:2.0 pytorch:2.0-distributed pytorch:2.1 pytorch:2.1-builder pytorch:2.1-distributed raft sam stable-diffusion stable-diffusion-webui tam text-generation-inference text-generation-webui:1.7 text-generation-webui:6a7cd01 text-generation-webui:main torch2trt torch_tensorrt torchaudio torchvision transformers transformers:git transformers:nvgpt tvm whisper whisperx xformers
Dockerfile	Dockerfile
Images	dustynv/onnx:r32.7.1 (2023-12-11, 0.4GB) dustynv/onnx:r35.2.1 (2023-12-12, 5.0GB) dustynv/onnx:r35.3.1 (2023-12-11, 5.0GB) dustynv/onnx:r35.4.1 (2023-10-07, 5.0GB)
Notes	protobuf_apt is added as a dependency on JetPack 4 (newer versions of onnx build it in-tree)

CONTAINER IMAGES

Repository/Tag	Date	Arch	Size
dustynv/onnx:r32.7.1	2023-12-11	arm64	0.4GB
dustynv/onnx:r35.2.1	2023-12-12	arm64	5.0GB
dustynv/onnx:r35.3.1	2023-12-11	arm64	5.0GB
dustynv/onnx:r35.4.1	2023-10-07	arm64	5.0GB

_{Container images are compatible with other minor versions of JetPack/L4T:
    • L4T R32.7 containers can run on other versions of L4T R32.7 (JetPack 4.6+)
    • L4T R35.x containers can run on other versions of L4T R35.x (JetPack 5.1+)}

RUN CONTAINER

To start the container, you can use the run.sh/autotag helpers or manually put together a docker run command:

# automatically pull or build a compatible container image
./run.sh $(./autotag onnx)

# or explicitly specify one of the container images above
./run.sh dustynv/onnx:r35.2.1

# or if using 'docker run' (specify image and mounts/ect)
sudo docker run --runtime nvidia -it --rm --network=host dustynv/onnx:r35.2.1

^{run.sh forwards arguments to docker run with some defaults added (like --runtime nvidia, mounts a /data cache, and detects devices)
autotag finds a container image that's compatible with your version of JetPack/L4T - either locally, pulled from a registry, or by building it.}

To mount your own directories into the container, use the -v or --volume flags:

./run.sh -v /path/on/host:/path/in/container $(./autotag onnx)

To launch the container running a command, as opposed to an interactive shell:

./run.sh $(./autotag onnx) my_app --abc xyz

You can pass any options to run.sh that you would to docker run, and it'll print out the full command that it constructs before executing it.

BUILD CONTAINER

If you use autotag as shown above, it'll ask to build the container for you if needed. To manually build it, first do the system setup, then run:

./build.sh onnx

The dependencies from above will be built into the container, and it'll be tested during. See ./build.sh --help for build options.