Skip to content

Latest commit

 

History

History
239 lines (173 loc) · 7.92 KB

installation.md

File metadata and controls

239 lines (173 loc) · 7.92 KB
Raw

Installation

Table of Contents

Installation with Pip

Install TensorFlow Decision Forests by running:

# Install TensorFlow Decision Forests.
pip3 install tensorflow_decision_forests --upgrade

Then, check the installation with by running:

# Check the version of TensorFlow Decision Forests.
python3 -c "import tensorflow_decision_forests as tfdf; print('Found TF-DF v' + tfdf.__version__)"

Note: Cuda warnings are not an issue.

Build from source

Linux

Setup

Requirements

  • Bazel >= 3.7.2
  • Python >= 3
  • Git
  • Python packages: numpy tensorflow pandas

Instead of installing the dependencies by hands, you can use the TensorFlow Build docker. If you choose this options, install Docker:

Compilation

Download TensorFlow Decision Forests as follows:

# Download the source code of TF-DF.
git clone https://github.com/tensorflow/decision-forests.git
cd decision-forests

Optional: TensorFlow Decision Forests depends on Yggdrasil Decision Forests . If you want to edit the Yggdrasil code, you can clone the Yggdrasil github and change the path accordingly in third_party/yggdrasil_decision_forests/workspace.bzl.

Optional: If you want to use the docker option, run the start_compile_docker.sh script and continue to the next step. If you don't want to use the docker option, continue to the next step directly.

# Optional: Install and start the build docker.
./tools/start_compile_docker.sh

Compile and run the unit tests of TF-DF with the following command. Note that test_bazel.sh is configured for python3.8 and the default compiler on your machine. Edit the file directly to change this configuration.

# Build and test TF-DF.
./tools/test_bazel.sh

Create and test a pip package with the following command. Replace python3.8 by the version of python you want to use. Note that you don't have to use the same version of Python as in the test_bazel.sh script.

If your configuration is compatible with manylinux2014, a manylinux2014 compatible pip package will be produced.

If your configuration is not compatible with manylinux2014, a non manylinux2014 compatible pip package will be produced, and the final check will fail. It does not matter if you want to use TF-DF on your own machine. An easy way to make the build manylinux2014 compatible is to use the docker mentioned above.

# Build and test a Pip package.
./tools/build_pip_package.sh python3.8

This command will install the TF-DF pip package and run the example in examples/minimal.py. The Pip package is located in the dist/ directory.

If you want to create a Pip package for the other compatible version of Python, run:

# Install the other versions of python (assume only python3.8 is installed; this is the case in the build docker).
sudo apt-get update && sudo apt-get install python3.7 python3.9 python3-pip

# Create the Pip package for the other version of python
./tools/build_pip_package.sh python3.7
./tools/build_pip_package.sh python3.9

Alternatively, you can create the pip package for all the compatible version of python using pyenv by running the following command. See the header of tools/build_pip_package.sh for more details.

# Build and test all the Pip package using Pyenv.
./tools/build_pip_package.sh ALL_VERSIONS

MacOS

Setup

Requirements

  • XCode command line tools
  • Bazel (recommended Bazelisk)
  • Python >= 3.8
  • Git
  • Pyenv (for building the Pip packages with multiple Python versions)

Building / Packaging (Apple CPU)

If you have a MacOS machine with Apple CPU, you can build with the following instructions.

  1. Clone the three repositories and adjust paths.

    git clone https://github.com/tensorflow/decision-forests.git
git clone https://github.com/google/yggdrasil-decision-forests.git
git clone --branch boost-1.75.0 https://github.com/boostorg/boost.git
(cd boost && git submodule update --init --checkout --force)
# Adjust path TF-DF --> YDF
perl -0777 -i.original -pe 's/    http_archive\(\n        name = "ydf",\n        urls = \["https:\/\/github.com\/google\/yggdrasil-decision-forests\/archive\/refs\/heads\/main.zip"\],\n        strip_prefix = "yggdrasil-decision-forests-main",\n    \)/    native.local_repository\(\n        name = "ydf",\n        path = "..\/yggdrasil-decision-forests",\n    \)/igs' decision-forests/third_party/yggdrasil_decision_forests/workspace.bzl
# Adjust path YDF --> Boost
perl -0777 -i.original -pe 's/    new_git_repository\(\n        name = "org_boost",\n        branch = branch,\n        build_file_content = build_file_content,\n        init_submodules = True,\n        recursive_init_submodules = True,\n        remote = "https:\/\/github.com\/boostorg\/boost",\n    \)/    native.new_local_repository\(\n        name = "org_boost",\n        path = "..\/boost",\n        build_file_content = build_file_content,\n    \)/igs' yggdrasil-decision-forests/third_party/boost/workspace.bzl

  2. (Optional) Create a fresh virtual environment and activate it

    python3 -m venv venv
source venv/source/activate

  3. Adjust the TensorFlow dependency for Apple CPUs

    perl -0777 -i.original -pe 's/tensorflow~=/tensorflow-macos~=/igs' decision-forests/configure/setup.py

  4. Decide which Python version you want to use and run

    cd decision-forests
# This will compile with the latest Tensorflow version in the tensorflow-macos repository.
RUN_TESTS=1 PY_VERSION=3.9 TF_VERSION=mac-arm64 ./tools/test_bazel.sh

  5. Build the Pip Packages

    # First, we deactivate our virtualenv, since the Pip script uses a different one.
deactivate
# Build the packages.
./tools/build_pip_package.sh ALL_VERSIONS_MAC_ARM64

  6. The packages can be found in decision-forests/dist/.

Cross-compiling for Intel CPUs

If you have a MacOS machine with Apple CPU, cross-compile TF-DF for MacOS machines with Intel CPUs as follows.

  1. Follow Steps 1-3 and 5 of the guide for Apple CPUs, skip Step 4.

  2. Decide which Python version you want to use and run

    cd decision-forests
# This will compile with the latest Tensorflow version in the tensorflow-macos repository.
RUN_TESTS=0 PY_VERSION=3.9 TF_VERSION=mac-intel-crosscompile ./tools/test_bazel.sh

  3. Build the Pip Packages

    # First, we deactivate our virtualenv, since the Pip script uses a different one.
deactivate
# Build the packages.
./tools/build_pip_package.sh ALL_VERSIONS_MAC_INTEL_CROSSCOMPILE

  4. The packages can be found in decision-forests/dist/.

Final note

Compiling TF-DF relies on the TensorFlow Pip package and the TensorFlow Bazel dependency. Only a small part of TensorFlow will be compiled. Compiling TF-DF on a single powerful workstation takes ~10 minutes.

Troubleshooting

Note: Check also Yggdrasil's Troubleshooting page.