You can get the latest source code from Github:
git clone https://github.com/sql-machine-learning/elasticdl
cd elasticdlWe prefer to install all building tools in a Docker image.
scripts/travis/build_images.shThe above Docker image contains pre-commit and hooks. We can run it as a container and mount the local Git repo into the container. In this container, we run the pre-commit command.
docker run --rm -it -v $PWD:/work -w /work elasticdl:dev \
bash -c "make -f elasticdl/Makefile && pre-commit run -a"If you have all required tools installed, you can also run the same script on your host.
make -f elasticdl/Makefile && pre-commit run -aAfter modifying code, you can build wheel packages by running the following command in the root of the project.
scripts/docker_build_wheel.shIf you have elasticdl:dev image, you can only run the following command to
generate wheel packages in the root of the project.
docker run --rm -it --net=host -v "$PWD":/work -w /work elasticdl:dev \
bash -c "scripts/build.sh"You can find wheel packages in the build directory.
After building wheel packages, you can build a docker image by
elasticdl zoo.
elasticdl zoo init \
--base_image=elasticdl:dev \
--model_zoo=model_zoo \
--local_pkg_dir=./build \
elasticdl zoo build --image=elasticdl:dev_test .Then, you can submit a job using the image like the tutorial.