TensorFlow Serving 源码编译安装

2019-06-11 22:41 阅读 734

系统环境：

linux系统，发行版 Ubuntu 12.04 LST

java8

Anancoda| Python3.6 -->tensorflow-gpu==1.4

1. 软件准备

为了可以编译并且使用Tesnsorflow Serving ，需要安装bazel和gRPC

1.1 Bazel安装

bazel是google的编译工具，目前tensorflow Serving要求的版本在0.5.4以上。

在github/bazel 中下载相应的版本，建议下载installer-linux-XXX.sh

chmod +x bazel-0.5.4-installer-linux-x86_64.sh

./bazel-0.5.4-installer-linux-x86_64.sh --user # --user是安装到当前用户的家目录下

在bashrc中添加path环境变量

export PATH="$PATH:$HOME/bin" #这里的$HOME是你的家目录

1.2 gRCP 安装

gRCP是google的RCP框架 [2][3]。gRCP的安装说明在这里here

gRCP依赖的软件如下：

sudo apt-get update && sudo apt-get install -y

build-essential

curl

libcurl3-dev

git

libfreetype6-dev

libpng12-dev

libzmq3-dev

pkg-config

software-properties-common

swig

zip

zlib1g-dev

接下来进入anaconda的root环境下安装即可：

pip install tensorflow-serving-api #grcp

你也可以用源代码编译的方式安装gRCP ,具体可以参看girhub的ReadME.md

2. TensorFlow Serving Installing from Source | 通过源代码的方式安装 Tensorflow Serving

2.1 下载Tensorflow serving

从github 下clone

git clone--recurse-submodules https://github.com/tensorflow/serving

#--recurse-submodules required to fetch TensorFlow, gRPC, and other libraries that TensorFlow Serving depends on. Note that these instructions will install the latest master branch of TensorFlow Serving. If you want to install a specific branch (such as a release branch), pass -b to the git clone command.

2.2 编译Tensorflow serving

下载完毕以后，将当前的工作路径切换到 serving/tensorflow中，进行配置

step1 configuration：

cd serving #切换到刚刚下载的目录中

cd tensorflow

./configure # 这里需要自己手动的操作了，有一点需要得别强调：

目前支持cuda8.0和cuda7.0

cudnn指定版本的时候，如果是下载cudann7.0.3, 请你输入7.0.3, 而不是7.0

cd .. # build完毕以后，返回上一级目录

step2 build

bazel build-c opt tensorflow_serving/...

如果编译成功的话，执行这个

bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server # 可以使用python bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server.py

为了验证效果，执行以下语句启动服务：

bazel test-c opt tensorflow_serving/...

好了，到此为止安装serving 的大致过程结束。

下期会介绍 basic tutorial and advanced tutorial 部分。

http://blog.csdn.net/dingqingsong/article/details/77931339

Optimized build

It's possible to compile using some platform specific instruction sets (e.g. AVX) that can significantly improve performance. Wherever you see 'bazel build' in the documentation, you can add the flags-c opt --copt=-msse4.1 --copt=-msse4.2 --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-O3(or some subset of these flags). For example:

bazel build-c opt--copt=-msse4.1--copt=-msse4.2--copt=-mavx--copt=-mavx2--copt=-mfma--copt=-O3 tensorflow_serving/...

Note:These instruction sets are not available on all machines, especially with older processors, so it may not work with all flags. You can try some subset of them, or revert to just the basic '-c opt' which is guaranteed to work on all machines.

Continuous integration build

Ourcontinuous integration buildusing TensorFlowci_buildinfrastructure offers you simplified development using docker. All you need is git and docker. No need to install all other dependencies manually.

git clone--recursive https://github.com/tensorflow/serving

cd serving

CI_TENSORFLOW_SUBMODULE_PATH

=tensorflow tensorflow/tensorflow/tools/ci_build/ci_build.sh CPU bazel test//tensorflow_serving/...

Note:Theservingdirectory is mapped into the container. You can develop outside the docker container (in your favourite editor) and when you run this build it will build with your changes.