TensorFlow for Java

Java bindings for TensorFlow.

WARNING: The TensorFlow Java API is incomplete and experimental and can change without notice. Progress can be followed in issue #5.

Till then, for using TensorFlow on Android refer to contrib/android, makefile and/or the Android camera demo.

Requirements

• bazel

• Environment to build TensorFlow from source code (Linux or Mac OS X). If you'd like to skip reading those details and do not care about GPU support, try the following:

  # On Linux
  sudo apt-get install python swig python-numpy
  
  # On Mac OS X with homebrew
  brew install swig
  

Installation

Build the Java Archive and native library:

bazel build -c opt \
  //tensorflow/java:libtensorflow.jar \
  //tensorflow/java:libtensorflow-jni

Example

With bazel

Add a dependency on //tensorflow/java:tensorflow to the java_binary or java_library rule. For example:

bazel run -c opt //tensorflow/java/src/main/java/org/tensorflow/examples:label_image

With javac

• Add libtensorflow.jar to classpath for compilation. For example:

  javac \
    -cp ../../bazel-bin/tensorflow/java/libtensorflow.jar \
    ./src/main/java/org/tensorflow/examples/LabelImage.java
  

• Make libtensorflow.jar and libtensorflow-jni.so (libtensorflow-jni.dylib on OS X) available during execution. For example:

  java \
    -Djava.library.path=../../bazel-bin/tensorflow/java \
    -cp ../../bazel-bin/tensorflow/java/libtensorflow.jar:./src/main/java \
    org.tensorflow.examples.LabelImage