Golang Install Guide

Documentation for installing the Go bindings for Tensorflow.

Maintainer: @wamuir

Important: TensorFlow for Go is no longer supported by the TensorFlow team.

Install TensorFlow for Go

TensorFlow provides a Go API particularly useful for loading models created with Python and running them within a Go application.

Caution: The TensorFlow Go API is not covered by the TensorFlow API stability guarantees.

Supported Platforms

The Go bindings for TensorFlow work on the following systems, and likely others:

• Linux, 64-bit, x86
• macOS, Version 10.12.6 (Sierra) or higher

Installation and Setup

1. Install the TensorFlow C library

Install the TensorFlow C library. This library is required for use of the TensorFlow Go package at runtime. For example, on Linux (64-bit, x86):

$ curl -L https://storage.googleapis.com/tensorflow/libtensorflow/libtensorflow-cpu-linux-x86_64-2.5.0.tar.gz | tar xz --directory /usr/local
$ ldconfig

2. Install the Protocol Buffer Compiler (Protoc)

Install the Protocol Buffer Compiler. This compiler is required to install the Go bindings but is not needed at runtime.

• Linux, using apt or apt-get, for example:

  $ apt install -y protobuf-compiler

• MacOS, using Homebrew:

  $ brew install protobuf

3. Install and Setup the Tensorflow Go API

The use of go get is not currently supported for installation of the Tensorflow Go API. Instead, follow these instructions.

• Decide on a location to install the API, such as within the src folder immediately below the Go workspace (e.g., the location of $GOPATH). The location /go/src/github.com/tensorflow/tensorflow will be used in these instructions.

• Clone the Tensorflow source respository to the install location

  $ git clone --branch v2.5.0 https://github.com/tensorflow/tensorflow.git /go/src/github.com/tensorflow/tensorflow

• Change the working directory to the install location.

  $ cd /go/src/github.com/tensorflow/tensorflow

• Apply a patch to declare the Go package within Tensorflow's proto definition files

  $ git format-patch -1 835d7da --stdout | git apply

• Initialize a new go.mod file

  $ go mod init github.com/tensorflow/tensorflow

• Generate the protocol buffers and move the generating files to their correct locations. You will receive two errors (no required module provides package ...), which you can ignore.

  $ cd tensorflow/go
  $ go generate ./...
  $ mv vendor/github.com/tensorflow/tensorflow/tensorflow/go/* .

• Add missing modules

  $ go mod tidy

• Test the installation

  $ go test ./...

Usage

Applications must use Go Mod's replace directive

The replace directive instructs Go to use the local installation. Point this to the installation location (such as /go/src/github.com/tensorflow/tensorflow from the above installation instructions).

This directive must be added for each Go module that uses the Go API. As an example:

$ go mod init hello-world
$ go mod edit -require github.com/tensorflow/tensorflow@v2.5.0+incompatible
$ go mod edit -replace github.com/tensorflow/tensorflow=/go/src/github.com/tensorflow/tensorflow
$ go mod tidy

Example program

With the TensorFlow Go package installed, create an example program with the following source code (hello_tf.go):

package main

import (
	tf "github.com/tensorflow/tensorflow/tensorflow/go"
	"github.com/tensorflow/tensorflow/tensorflow/go/op"
	"fmt"
)

func main() {
	// Construct a graph with an operation that produces a string constant.
	s := op.NewScope()
	c := op.Const(s, "Hello from TensorFlow version " + tf.Version())
	graph, err := s.Finalize()
	if err != nil {
		panic(err)
	}

	// Execute the graph in a session.
	sess, err := tf.NewSession(graph, nil)
	if err != nil {
		panic(err)
	}
	output, err := sess.Run(nil, []tf.Output{c}, nil)
	if err != nil {
		panic(err)
	}
	fmt.Println(output[0].Value())
}

Initialize go.mod for the example program:

$ go mod init app
$ go mod edit -require github.com/tensorflow/tensorflow@v2.5.0+incompatible
$ go mod edit -replace github.com/tensorflow/tensorflow=/go/src/github.com/tensorflow/tensorflow
$ go mod tidy

Then, run the example program:

$ go run hello_tf.go

The command outputs: Hello from TensorFlow version *number*

Success: TensorFlow for Go has been configured.

The program may generate the following warning messages, which you can ignore:

W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library
wasn't compiled to use *Type* instructions, but these are available on your
machine and could speed up CPU computations.

Docker Example

A Dockerfile is available, which executes the installation and setup process for the Go bindings and builds the example program. To use, install Docker and then run the following commands:

$ docker build -t tensorflow/build:golang-example https://github.com/tensorflow/build.git#:golang_install_guide/example-program
$ docker run -it --rm tensorflow/build:golang-example

Build from source

TensorFlow is open source. Read the instructions to build TensorFlow for Go from source code.