Git Product home page Git Product logo

caffe's Introduction

Introduction

This is my repository to study Caffe, which is a great deep learning framework.There are some annotations in this work to clarify the code of framework.I think that the programer can improve a lot from the code of Caffe.

caffe是深度学习领域优秀的框架,深入其源码有助于更好地应用caffe,更能够通过阅读优秀的代码实践,来提升自己的编码能力,同时也能够加深对于算法的理解。

在这个工程中,我将记录我研读caffe源代码的一些心得和注解。

installation

caffe的安装方法,有很多教程,官网上也给出了安装方法。这里我给出一些建议,推荐在ubuntu系统上进行安装,这样会更简单,不会遇到太多复杂的问题。

1.安装依赖
caffe运行依赖于很多第三方的库,这些依赖需要提前安装

# git
sudo apt-get install git

# protobuffer,leveldb,snappy,opencv,hdf5
sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler

# boost library
sudo apt-get install --no-install-recommends libboost-all-dev

# atlas library
sudo apt-get install libatlas-base-dev

# python dev
sudo apt-get install python-dev

# flags,glog,lmdb
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev

2.安装caffe
建议使用cmake对caffe进行编译安装。在多次安装caffe的经验中,这样是最简单的且不容易出错。

# in the directory of caffe
mkdir build
cd build

# cmake
cmake ..

# compile and install
make all
make install
make runtest

详情请参考:http://caffe.berkeleyvision.org/installation.html

protobuffer

protobuffer是由谷歌研发的对象序列化和反序列化的开源工具,是一种数据描述的工具,其功能类似于xml、json。但是protobuffer最终的数据描述文件是二进制的格式,这就造成了他阅读不直观,但是占用空间小,解析传输速度更快。

首先编写描述定义文件proto,再通过protobuffer的编译器protoc对描述文件进行编译,得到可直接使用的二进制文件(支持C++、python、java)。例如,在C++代码中,会直接将demo.proto文件编译成可供代码直接使用的demo.pb.h和demo.pb.cc。

在caffe中,blob定义、网络定义等都使用了protobuffer,使用者通过编写网络定义proto文件控制caffe构建网络。

详情请参考:https://developers.google.cn/protocol-buffers/

blob

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.