diff options
Diffstat (limited to 'tensorflow/tools/docker/README.md')
| -rw-r--r-- | tensorflow/tools/docker/README.md | 28 |
1 files changed, 2 insertions, 26 deletions
diff --git a/tensorflow/tools/docker/README.md b/tensorflow/tools/docker/README.md index 6d5a9bdc4c..3e45ae362c 100644 --- a/tensorflow/tools/docker/README.md +++ b/tensorflow/tools/docker/README.md @@ -54,30 +54,6 @@ for additional containers, such as release candidates or nightly builds. ## Rebuilding the containers -Building TensorFlow Docker containers should be done through the -[parameterized_docker_build.sh](https://github.com/tensorflow/tensorflow/blob/master/tensorflow/tools/docker/README.md) -script. The raw Dockerfiles should not be used directly as they contain strings -to be replaced by the script during the build. +Just pick the dockerfile corresponding to the container you want to build, and run -To use the script, specify the container type (`CPU` vs. `GPU`), the desired -Python version (`PYTHON2` vs. `PYTHON3`) and whether the developer Docker image -is to be built (`NO` vs. `YES`). In addition, you need to specify the central -location from where the pip package of TensorFlow will be downloaded. - -For example, to build a CPU-only non-developer Docker image for Python 2, using -TensorFlow's nightly pip package: - -``` bash -export TF_DOCKER_BUILD_IS_DEVEL=NO -export TF_DOCKER_BUILD_TYPE=CPU -export TF_DOCKER_BUILD_PYTHON_VERSION=PYTHON2 - -export NIGHTLY_VERSION="1.head" -export TF_DOCKER_BUILD_CENTRAL_PIP=$(echo ${TF_DOCKER_BUILD_PYTHON_VERSION} | sed s^PYTHON2^http://ci.tensorflow.org/view/Nightly/job/nightly-matrix-cpu/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=${TF_DOCKER_BUILD_PYTHON_VERSION},label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-${NIGHTLY_VERSION}-cp27-cp27mu-manylinux1_x86_64.whl^ | sed s^PYTHON3^http://ci.tensorflow.org/view/Nightly/job/nightly-python35-linux-cpu/lastSuccessfulBuild/artifact/pip_test/whl/tensorflow-${NIGHTLY_VERSION}-cp35-cp35m-manylinux1_x86_64.whl^) - -tensorflow/tools/docker/parameterized_docker_build.sh -``` - -If successful, the image will be tagged as `${USER}/tensorflow:latest` by default. - -Rebuilding GPU images requires [nvidia-docker](https://github.com/NVIDIA/nvidia-docker). + $ docker build --pull -t $USER/tensorflow-suffix -f Dockerfile.suffix . |