derekmerck/diana repository overview

⁠DIANA xArch Docker Image

Derek Merck
[email protected]⁠
Rhode Island Hospital and Brown University
Providence, RI

Build multi-arch DIANA⁠ and DIANA-Learn Python Docker images for embedded systems.

⁠Use It

$ docker run derekmerck/diana:latest    # (amd64, arm32v7, arm64v8)
$ docker run derekmerck/diana:learn     # (learn-amd64, learn-arm32v7, learn-arm64v8)
$ docker run derekmerck/diana:movidius  # (movidius-arm32v7)

⁠Build It

This image uses system python and the resin/$ARCH-debian:stretch image. Resin.io⁠ base images include a QEMU⁠ cross-compiler to facilitate building images for low-power single-board computers on more powerful Intel-architecture desktops and servers.

docker-compose.yml contains build descriptions for all relevant architectures.

⁠amd64

$ docker-compose build diana-amd64

Desktop computers/vms, UP boards⁠, and the Intel NUC⁠ are amd64 devices. The appropriate image can be built and pushed from Travis CI⁠.

⁠arm32v7

Most low-power single board computers such as the Raspberry Pi and Beagleboard are arm32v7 devices. Appropriate images can be cross-compiled and pushed from Travis CI.

$ docker-compose build diana-arm32v7 diana-movidius-arm32v7

The official arm32 tensorflow wheels are available from pypi or as nightly build artifacts⁠. The wheel name for the python 3.4 build has to be manipuated to remove the platform restriction tags in order to install on 3.5 or 3.6.

The Intel Movidius⁠ NPU drivers from the NCSDK v2.0⁠ are available in the diana:movidius tag. Only the toolkit itself is installed, tensorflow is from pypi and caffe⁠ must be installed separately if needed.

⁠arm64v8

The NVIDIA Jetson TX2⁠ uses a Tegra arm64v8 cpu. The appropriate image can be built natively and pushed from Packet.io⁠, using a brief tenancy on a bare-metal Cavium ThunderX ARMv8 server.

$ apt update && apt upgrade
$ curl -fsSL get.docker.com -o get-docker.sh
$ sh get-docker.sh 
$ docker run hello-world
$ apt install git python-pip
$ pip install docker-compose
$ git clone http://github.com/derekmerck/diana-xarch@system_python
$ cd orthanc-xarch
$ docker-compose build diana-arm64v8
$ docker-compose build diana-learn-arm64v8
$ python3 manifest-it.py diana-xarch.manifest.yml

Although Resin uses Packet ARM servers to compile arm32 images⁠, the available ThunderX does not implement the arm32 instruction set, so it cannot compile natively for the Raspberry Pi⁠.

NVIDIA provides a recent tensorflow wheel for their Jetson TXs⁠.

⁠Manifest It

After building new images, call manifest-it.py to push updated images and build the Docker multi-architecture service mappings.

$ python3 manifest-it diana-xarch.manifest.yml

⁠License

MIT