sophon-ai-algo / docker-images

English | 中文

Building the conversion docker image

Introduction

The conversion docker image's Dockerfile corresponds to docker-images/Dockerfile/convert/ubuntu1604/Dockerfile.*.

There are currently two Dockerfile based on ubuntu1604, the differences of which are described later.

Preparation

• Install Docker

  The following Dockerfile can be built successfully on Docker version 20.10.7

• High-quality network

Dockerfile description

• Dockerfile.all contains Python3.5 3.6 3.7 3.8 Python2.7, where Python3 is managed using update-alternatives and Python3.6 3.7 3.8 requires installing pip manually .

  Use update-alternatives --config python3 to switch between versions
  After successful switch, use curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py && python3 get-pip.py to install pip

• Dockerfile.select contains Python2.7 Python3.7 Python3.5 by default, where Python3.7 can be replaced by Python3.6 Python3.8 at the build stage of the image, and the user does not need to install pip.

Building images

• If you choose Dockerfile.select to build the image, you need to specify the Python3 version with the PYTHON_VERSION parameter, otherwise Python3.7 will be installed by default.

  docker build -t bmnnsdk2-bm1684/dev:ubuntu1604 -f Dockerfile.select --build-arg PYTHON_VERSION=3.x
3.x can be replaced with 3.6 3.7 3.8

• If the user chooses Dockerfile.all for image building

  docker build -t bmnnsdk2-bm1684/dev:ubuntu1604 -f Dockerfile.all
  Note that using this Dockerfile will multiple times compile the Python source code, so please be patient.

Use

Please use it with the SDK, see the tutorial on the official website for details.

Building the deployment image

Introduction

The deployment image's Dockerfile corresponds to docker-images/Dockerfile/deploy/x86/ubuntu1604/Dockerfile/*.

Preparation

• Install Docker

  The following Dockerfile can be built successfully on Docker version 20.10.7

• High-quality network

Dockerfile description

• If you choose Dockerfile.select to build the image, you need to specify the Python3 version with the PYTHON_VERSION parameter, otherwise Python3.7 will be installed by default.

  docker build -t bmnnsdk2-bm1684/product:ubuntu1604 -f Dockerfile.select --build-arg PYTHON_VERSION=3.x
3.x can be replaced with 3.6 3.7 3.8

• If the user chooses Dockerfile.all for image building

  docker build -t bmnnsdk2-bm1684/product:ubuntu1604 -f Dockerfile.all
  Note that using this Dockerfile will multiple times compile the Python source code, so please be patient.

Use

with the full SDK

1. Go to the SDK root directory.
2. Create the script file start_docker.sh, the content of which is referenced below and can be adapted to suit your needs.

DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
WORKSPACE=$PWD
if [ -c "/dev/bm-sophon0" ]; then
  for dev in $(ls /dev/bm-sophon*);
  do
    mount_options+="--device="$dev:$dev" "
  done
fi
docker run \
      --workdir=/workspace \
      --privileged=true \
      ${mount_options} \
      --device=/dev/bmdev-ctl:/dev/bmdev-ctl \
      -v /dev/shm --tmpfs /dev/shm:exec \
      -v $WORKSPACE:/workspace \
      -v /dev:/dev \
      -v /etc/localtime:/etc/localtime \
      -e LOCAL_USER_ID=`id -u` \
      -it bmnnsdk2-bm1684/product:ubuntu1604 \
      bash

3. Run the script to start Docker

4. cd scripts && ./install_lib.sh nntc && source ./envsetup_pcie.sh