梳理caffe代码common(八)
来源:互联网 发布:周润发占中言论 知乎 编辑:程序博客网 时间:2024/06/06 11:42
由于想梳理data_layer的过程,整理一半发现有几个非常重要的头文件就是题目列出的这几个:
追本溯源,先从根基开始学起。这里面都是些什么鬼呢?
common类
命名空间的使用:google、cv、caffe{boost、std}。然后在项目中就可以随意使用google、opencv、c++的标准库、以及c++高级库boost。caffe采用单例模式封装boost的智能指针(caffe的灵魂)、std一些标准的用法、重要的初始化内容(随机数生成器的内容以及google的gflags和glog的初始化)。 提供一个统一的接口,方便移植和开发。为毛使用随机数?我也不是很清楚,知乎的一个解释:
随机数在caffe中是非常重要的,最重要的应用是权值的初始化,如高斯、xavier等,初始化的好坏直接影响最终的训练结果,其他的应用如训练图像的随机crop和mirror、dropout层的神经元的选择。RNG类是对Boost以及STL中随机数函数的封装,以方便使用。至于想每次产生相同的随机数,只要设定固定的种子即可,见caffe.proto中random_seed的定义:
// If non-negative, the seed with which the Solver will initialize the Caffe
// random number generator -- useful for reproducible results. Otherwise,
// (and by default) initialize using a seed derived from the system clock.
optional int64 random_seed = 20 [default = -1];
头文件:
#ifndef CAFFE_COMMON_HPP_#define CAFFE_COMMON_HPP_#include <boost/shared_ptr.hpp>#include <gflags/gflags.h>#include <glog/logging.h>#include <climits>#include <cmath>#include <fstream> // NOLINT(readability/streams)#include <iostream> // NOLINT(readability/streams)#include <map>#include <set>#include <sstream>#include <string>#include <utility> // pair#include <vector>#include "caffe/util/device_alternate.hpp"// Convert macro to string// 将宏转换为字符串#define STRINGIFY(m) #m#define AS_STRING(m) STRINGIFY(m)// gflags 2.1 issue: namespace google was changed to gflags without warning.// Luckily we will be able to use GFLAGS_GFLAGS_H_ to detect if it is version// 2.1. If yes, we will add a temporary solution to redirect the namespace.// TODO(Yangqing): Once gflags solves the problem in a more elegant way, let's// remove the following hack.// 检测gflags2.1#ifndef GFLAGS_GFLAGS_H_namespace gflags = google;#endif // GFLAGS_GFLAGS_H_// Disable the copy and assignment operator for a class.// 禁止某个类通过构造函数直接初始化另一个类// 禁止某个类通过赋值来初始化另一个类#define DISABLE_COPY_AND_ASSIGN(classname) \private:\ classname(const classname&);\ classname& operator=(const classname&)// Instantiate a class with float and double specifications.#define INSTANTIATE_CLASS(classname) \ char gInstantiationGuard##classname; \ template class classname<float>; \ template class classname<double>// 初始化GPU的前向传播函数#define INSTANTIATE_LAYER_GPU_FORWARD(classname) \ template void classname<float>::Forward_gpu( \ const std::vector<Blob<float>*>& bottom, \ const std::vector<Blob<float>*>& top); \ template void classname<double>::Forward_gpu( \ const std::vector<Blob<double>*>& bottom, \ const std::vector<Blob<double>*>& top);// 初始化GPU的反向传播函数#define INSTANTIATE_LAYER_GPU_BACKWARD(classname) \ template void classname<float>::Backward_gpu( \ const std::vector<Blob<float>*>& top, \ const std::vector<bool>& propagate_down, \ const std::vector<Blob<float>*>& bottom); \ template void classname<double>::Backward_gpu( \ const std::vector<Blob<double>*>& top, \ const std::vector<bool>& propagate_down, \ const std::vector<Blob<double>*>& bottom)// 初始化GPU的前向反向传播函数#define INSTANTIATE_LAYER_GPU_FUNCS(classname) \ INSTANTIATE_LAYER_GPU_FORWARD(classname); \ INSTANTIATE_LAYER_GPU_BACKWARD(classname)// A simple macro to mark codes that are not implemented, so that when the code// is executed we will see a fatal log.// NOT_IMPLEMENTED实际上调用的LOG(FATAL) << "Not Implemented Yet"#define NOT_IMPLEMENTED LOG(FATAL) << "Not Implemented Yet"// See PR #1236namespace cv { class Mat; }/*Caffe类里面有个RNG,RNG这个类里面还有个Generator类在RNG里面会用到Caffe里面的Get()函数来获取一个新的Caffe类的实例。然后RNG里面用到了Generator。Generator是实际产生随机数的。*/namespace caffe {// We will use the boost shared_ptr instead of the new C++11 one mainly// because cuda does not work (at least now) well with C++11 features.using boost::shared_ptr;// Common functions and classes from std that caffe often uses.using std::fstream;using std::ios;//using std::isnan;//vc++的编译器不支持这两个函数//using std::isinf;using std::iterator;using std::make_pair;using std::map;using std::ostringstream;using std::pair;using std::set;using std::string;using std::stringstream;using std::vector;// A global initialization function that you should call in your main function.// Currently it initializes google flags and google logging.void GlobalInit(int* pargc, char*** pargv);// A singleton class to hold common caffe stuff, such as the handler that// caffe is going to use for cublas, curand, etc.class Caffe { public: ~Caffe(); // Thread local context for Caffe. Moved to common.cpp instead of // including boost/thread.hpp to avoid a boost/NVCC issues (#1009, #1010) // on OSX. Also fails on Linux with CUDA 7.0.18.//Get函数利用Boost的局部线程存储功能实现 static Caffe& Get();//Brew就是CPU,GPU的枚举类型,这个名字是不是来自Homebrew???Mac的软件包管理器,我猜的。。。。 enum Brew { CPU, GPU }; // This random number generator facade hides boost and CUDA rng // implementation from one another (for cross-platform compatibility). class RNG { public: RNG();//利用系统的熵池或者时间来初始化RNG内部的generator_ explicit RNG(unsigned int seed); explicit RNG(const RNG&); RNG& operator=(const RNG&); void* generator(); private: class Generator; shared_ptr<Generator> generator_; }; // Getters for boost rng, curand, and cublas handles inline static RNG& rng_stream() { if (!Get().random_generator_) { Get().random_generator_.reset(new RNG()); } return *(Get().random_generator_); }#ifndef CPU_ONLY// GPU inline static cublasHandle_t cublas_handle() { return Get().cublas_handle_; }// cublas的句柄 inline static curandGenerator_t curand_generator() {//curandGenerator句柄 return Get().curand_generator_; }#endif//下面这一块就是设置CPU和GPU以及训练的时候线程并行数目吧 // Returns the mode: running on CPU or GPU. inline static Brew mode() { return Get().mode_; } // The setters for the variables // Sets the mode. It is recommended that you don't change the mode halfway // into the program since that may cause allocation of pinned memory being // freed in a non-pinned way, which may cause problems - I haven't verified // it personally but better to note it here in the header file. inline static void set_mode(Brew mode) { Get().mode_ = mode; } // Sets the random seed of both boost and curand static void set_random_seed(const unsigned int seed); // Sets the device. Since we have cublas and curand stuff, set device also // requires us to reset those values. static void SetDevice(const int device_id); // Prints the current GPU status. static void DeviceQuery(); // Parallel training info inline static int solver_count() { return Get().solver_count_; } inline static void set_solver_count(int val) { Get().solver_count_ = val; } inline static bool root_solver() { return Get().root_solver_; } inline static void set_root_solver(bool val) { Get().root_solver_ = val; } protected:#ifndef CPU_ONLY cublasHandle_t cublas_handle_;// cublas的句柄 curandGenerator_t curand_generator_;// curandGenerator句柄#endif shared_ptr<RNG> random_generator_; Brew mode_; int solver_count_; bool root_solver_; private: // The private constructor to avoid duplicate instantiation.//避免实例化 Caffe(); // 禁止caffe这个类被复制构造函数和赋值进行构造 DISABLE_COPY_AND_ASSIGN(Caffe);};} // namespace caffe#endif // CAFFE_COMMON_HPP_cpp文件:
#include <boost/thread.hpp>#include <glog/logging.h>#include <cmath>#include <cstdio>#include <ctime>#include "caffe/common.hpp"#include "caffe/util/rng.hpp"namespace caffe {// Make sure each thread can have different values.// boost::thread_specific_ptr是线程局部存储机制// 一开始的值是NULLstatic boost::thread_specific_ptr<Caffe> thread_instance_;Caffe& Caffe::Get() { if (!thread_instance_.get()) {// 如果当前线程没有caffe实例 thread_instance_.reset(new Caffe());// 则新建一个caffe的实例并返回 } return *(thread_instance_.get());}// random seeding// linux下的熵池下获取随机数的种子int64_t cluster_seedgen(void) { int64_t s, seed, pid; FILE* f = fopen("/dev/urandom", "rb"); if (f && fread(&seed, 1, sizeof(seed), f) == sizeof(seed)) { fclose(f); return seed; } LOG(INFO) << "System entropy source not available, " "using fallback algorithm to generate seed instead."; if (f) fclose(f); // 采用传统的基于时间来生成随机数种子 pid = getpid(); s = time(NULL); seed = std::abs(((s * 181) * ((pid - 83) * 359)) % 104729); return seed;}// 初始化gflags和glogvoid GlobalInit(int* pargc, char*** pargv) { // Google flags. ::gflags::ParseCommandLineFlags(pargc, pargv, true); // Google logging. ::google::InitGoogleLogging(*(pargv)[0]); // Provide a backtrace on segfault. ::google::InstallFailureSignalHandler();}#ifdef CPU_ONLY // CPU-only Caffe.Caffe::Caffe() : random_generator_(), mode_(Caffe::CPU),// shared_ptr<RNG> random_generator_; Brew mode_; solver_count_(1), root_solver_(true) { }// int solver_count_; bool root_solver_;Caffe::~Caffe() { }// 手动设定随机数生成器的种子void Caffe::set_random_seed(const unsigned int seed) { // RNG seed Get().random_generator_.reset(new RNG(seed));<span style="font-family:Microsoft YaHei;">}</span>void Caffe::SetDevice(const int device_id) { NO_GPU;}void Caffe::DeviceQuery() { NO_GPU;}// 定义RNG内部的Generator类class Caffe::RNG::Generator { public: Generator() : rng_(new caffe::rng_t(cluster_seedgen())) {}// linux下的熵池生成随机数种子,注意typedef boost::mt19937 rng_t;这个在utils/rng.hpp头文件里面 explicit Generator(unsigned int seed) : rng_(new caffe::rng_t(seed)) {}// 采用给定的种子初始化 caffe::rng_t* rng() { return rng_.get(); }// 属性 private: shared_ptr<caffe::rng_t> rng_;// 内部变量};// 实现RNG内部的构造函数Caffe::RNG::RNG() : generator_(new Generator()) { }Caffe::RNG::RNG(unsigned int seed) : generator_(new Generator(seed)) { }// 实现RNG内部的运算符重载Caffe::RNG& Caffe::RNG::operator=(const RNG& other) { generator_ = other.generator_; return *this;}void* Caffe::RNG::generator() { return static_cast<void*>(generator_->rng());}#else // Normal GPU + CPU Caffe.// 构造函数,初始化cublas和curand库的句柄Caffe::Caffe() : cublas_handle_(NULL), curand_generator_(NULL), random_generator_(), mode_(Caffe::CPU), solver_count_(1), root_solver_(true) { // Try to create a cublas handler, and report an error if failed (but we will // keep the program running as one might just want to run CPU code). // 初始化cublas并获得句柄 if (cublasCreate(&cublas_handle_) != CUBLAS_STATUS_SUCCESS) { LOG(ERROR) << "Cannot create Cublas handle. Cublas won't be available."; } // Try to create a curand handler. if (curandCreateGenerator(&curand_generator_, CURAND_RNG_PSEUDO_DEFAULT) != CURAND_STATUS_SUCCESS || curandSetPseudoRandomGeneratorSeed(curand_generator_, cluster_seedgen()) != CURAND_STATUS_SUCCESS) { LOG(ERROR) << "Cannot create Curand generator. Curand won't be available."; }}Caffe::~Caffe() { // 销毁句柄 if (cublas_handle_) CUBLAS_CHECK(cublasDestroy(cublas_handle_)); if (curand_generator_) { CURAND_CHECK(curandDestroyGenerator(curand_generator_)); }}// 初始化CUDA的随机数种子以及cpu的随机数种子void Caffe::set_random_seed(const unsigned int seed) { // Curand seed static bool g_curand_availability_logged = false;// 判断是否log了curand的可用性,如果没有则log一次,log之后则再也不log,用的是静态变量 if (Get().curand_generator_) { // CURAND_CHECK见/utils/device_alternate.hpp中的宏定义 CURAND_CHECK(curandSetPseudoRandomGeneratorSeed(curand_generator(), seed)); CURAND_CHECK(curandSetGeneratorOffset(curand_generator(), 0)); } else { if (!g_curand_availability_logged) { LOG(ERROR) << "Curand not available. Skipping setting the curand seed."; g_curand_availability_logged = true; } } // RNG seed // CPU code Get().random_generator_.reset(new RNG(seed));}// 设置GPU设备并初始化句柄以及随机数种子void Caffe::SetDevice(const int device_id) { int current_device; CUDA_CHECK(cudaGetDevice(¤t_device));// 获取当前设备id if (current_device == device_id) { return; } // The call to cudaSetDevice must come before any calls to Get, which // may perform initialization using the GPU. // 在Get之前必须先执行cudasetDevice函数 CUDA_CHECK(cudaSetDevice(device_id)); // 清理以前的句柄 if (Get().cublas_handle_) CUBLAS_CHECK(cublasDestroy(Get().cublas_handle_)); if (Get().curand_generator_) { CURAND_CHECK(curandDestroyGenerator(Get().curand_generator_)); } // 创建新句柄 CUBLAS_CHECK(cublasCreate(&Get().cublas_handle_)); CURAND_CHECK(curandCreateGenerator(&Get().curand_generator_, CURAND_RNG_PSEUDO_DEFAULT)); // 设置随机数种子 CURAND_CHECK(curandSetPseudoRandomGeneratorSeed(Get().curand_generator_, cluster_seedgen()));}// 获取设备信息void Caffe::DeviceQuery() { cudaDeviceProp prop; int device; if (cudaSuccess != cudaGetDevice(&device)) { printf("No cuda device present.\n"); return; } // #define CUDA_CHECK(condition) \ /* Code block avoids redefinition of cudaError_t error */ \ //do { \ // cudaError_t error = condition; \ // CHECK_EQ(error, cudaSuccess) << " " << cudaGetErrorString(error); \ //} while (0) CUDA_CHECK(cudaGetDeviceProperties(&prop, device)); LOG(INFO) << "Device id: " << device; LOG(INFO) << "Major revision number: " << prop.major; LOG(INFO) << "Minor revision number: " << prop.minor; LOG(INFO) << "Name: " << prop.name; LOG(INFO) << "Total global memory: " << prop.totalGlobalMem; LOG(INFO) << "Total shared memory per block: " << prop.sharedMemPerBlock; LOG(INFO) << "Total registers per block: " << prop.regsPerBlock; LOG(INFO) << "Warp size: " << prop.warpSize; LOG(INFO) << "Maximum memory pitch: " << prop.memPitch; LOG(INFO) << "Maximum threads per block: " << prop.maxThreadsPerBlock; LOG(INFO) << "Maximum dimension of block: " << prop.maxThreadsDim[0] << ", " << prop.maxThreadsDim[1] << ", " << prop.maxThreadsDim[2]; LOG(INFO) << "Maximum dimension of grid: " << prop.maxGridSize[0] << ", " << prop.maxGridSize[1] << ", " << prop.maxGridSize[2]; LOG(INFO) << "Clock rate: " << prop.clockRate; LOG(INFO) << "Total constant memory: " << prop.totalConstMem; LOG(INFO) << "Texture alignment: " << prop.textureAlignment; LOG(INFO) << "Concurrent copy and execution: " << (prop.deviceOverlap ? "Yes" : "No"); LOG(INFO) << "Number of multiprocessors: " << prop.multiProcessorCount; LOG(INFO) << "Kernel execution timeout: " << (prop.kernelExecTimeoutEnabled ? "Yes" : "No"); return;}class Caffe::RNG::Generator { public: Generator() : rng_(new caffe::rng_t(cluster_seedgen())) {} explicit Generator(unsigned int seed) : rng_(new caffe::rng_t(seed)) {} caffe::rng_t* rng() { return rng_.get(); } private: shared_ptr<caffe::rng_t> rng_;};Caffe::RNG::RNG() : generator_(new Generator()) { }Caffe::RNG::RNG(unsigned int seed) : generator_(new Generator(seed)) { }Caffe::RNG& Caffe::RNG::operator=(const RNG& other) { generator_.reset(other.generator_.get()); return *this;}void* Caffe::RNG::generator() { return static_cast<void*>(generator_->rng());}// cublas的geterrorstringconst char* cublasGetErrorString(cublasStatus_t error) { switch (error) { case CUBLAS_STATUS_SUCCESS: return "CUBLAS_STATUS_SUCCESS"; case CUBLAS_STATUS_NOT_INITIALIZED: return "CUBLAS_STATUS_NOT_INITIALIZED"; case CUBLAS_STATUS_ALLOC_FAILED: return "CUBLAS_STATUS_ALLOC_FAILED"; case CUBLAS_STATUS_INVALID_VALUE: return "CUBLAS_STATUS_INVALID_VALUE"; case CUBLAS_STATUS_ARCH_MISMATCH: return "CUBLAS_STATUS_ARCH_MISMATCH"; case CUBLAS_STATUS_MAPPING_ERROR: return "CUBLAS_STATUS_MAPPING_ERROR"; case CUBLAS_STATUS_EXECUTION_FAILED: return "CUBLAS_STATUS_EXECUTION_FAILED"; case CUBLAS_STATUS_INTERNAL_ERROR: return "CUBLAS_STATUS_INTERNAL_ERROR";#if CUDA_VERSION >= 6000 case CUBLAS_STATUS_NOT_SUPPORTED: return "CUBLAS_STATUS_NOT_SUPPORTED";#endif#if CUDA_VERSION >= 6050 case CUBLAS_STATUS_LICENSE_ERROR: return "CUBLAS_STATUS_LICENSE_ERROR";#endif } return "Unknown cublas status";}// curand的getlasterrorstringconst char* curandGetErrorString(curandStatus_t error) { switch (error) { case CURAND_STATUS_SUCCESS: return "CURAND_STATUS_SUCCESS"; case CURAND_STATUS_VERSION_MISMATCH: return "CURAND_STATUS_VERSION_MISMATCH"; case CURAND_STATUS_NOT_INITIALIZED: return "CURAND_STATUS_NOT_INITIALIZED"; case CURAND_STATUS_ALLOCATION_FAILED: return "CURAND_STATUS_ALLOCATION_FAILED"; case CURAND_STATUS_TYPE_ERROR: return "CURAND_STATUS_TYPE_ERROR"; case CURAND_STATUS_OUT_OF_RANGE: return "CURAND_STATUS_OUT_OF_RANGE"; case CURAND_STATUS_LENGTH_NOT_MULTIPLE: return "CURAND_STATUS_LENGTH_NOT_MULTIPLE"; case CURAND_STATUS_DOUBLE_PRECISION_REQUIRED: return "CURAND_STATUS_DOUBLE_PRECISION_REQUIRED"; case CURAND_STATUS_LAUNCH_FAILURE: return "CURAND_STATUS_LAUNCH_FAILURE"; case CURAND_STATUS_PREEXISTING_FAILURE: return "CURAND_STATUS_PREEXISTING_FAILURE"; case CURAND_STATUS_INITIALIZATION_FAILED: return "CURAND_STATUS_INITIALIZATION_FAILED"; case CURAND_STATUS_ARCH_MISMATCH: return "CURAND_STATUS_ARCH_MISMATCH"; case CURAND_STATUS_INTERNAL_ERROR: return "CURAND_STATUS_INTERNAL_ERROR"; } return "Unknown curand status";}#endif // CPU_ONLY} // namespace caffe
- 梳理caffe代码common(八)
- Caffe代码梳理笔记
- 梳理caffe代码layer_factory
- 梳理caffe代码math_functions
- 梳理caffe代码math_functions(一)
- 梳理caffe代码syncedmem(二)
- 梳理caffe代码blob(三)
- 梳理caffe代码net(四)
- 梳理caffe代码layer(五)
- 梳理caffe代码layer_factory(六)
- 梳理caffe代码python_layer(十五)
- 梳理caffe代码layer(五)
- caffe代码layer_factory梳理分析
- 梳理caffe代码math_functions(一)
- 梳理caffe代码image_data_layer、data_layer、window_data_layer(七)
- 梳理caffe代码internal_thread(九)
- 梳理caffe代码blocking_queue(十)
- 梳理caffe代码data_reader(十一)
- 4种Java引用浅解
- autocomplete 插件用法和用js实现autocomplete
- Java线程同步-牛逼的线程同步实现
- 备忘录模式(Memento)
- vector用法入门
- 梳理caffe代码common(八)
- Android Vibrator类
- 时钟周期、振荡周期、机器周期、CPU周期、状态周期、指令周期、总线周期、任务周期简单介绍。
- CAS 4.0
- 字符转转十六进制,再转十进制
- 归并排序-迭代法与递归法
- 在 iTunes Store 中支付未付余款
- 虚拟机隔离
- 替换wsock32.dl