word2vec之源码注释

2015-03-20 By SuHui

//  main.c
//  word2vec
//
//  Created by suhui on 15/3/20.
//  Copyright (c) 2015年 suhui. All rights reserved.
//

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <math.h> 
#include <pthread.h> 

// 一个word的最大长度
#define MAX_STRING 100
// 对f的运算结果进行缓存,存储1000个,需要用的时候查表
#define EXP_TABLE_SIZE 1000
// 最大计算到6 (exp^6 / (exp^6 + 1)),最小计算到-6 (exp^-6 / (exp^-6 + 1))
#define MAX_EXP 6
// 定义最大的句子长度
#define MAX_SENTENCE_LENGTH 1000
// 定义最长的霍夫曼编码长度
#define MAX_CODE_LENGTH 40
// 哈希,线性探测,开放定址法,装填系数0.7
const int vocab_hash_size = 30000000;

//重命名浮点数
typedef float real;

struct vocab_word {
    long long cn; //单词词频
    int * point ; //霍夫曼树中从根节点到该词的路径,存放路径上每个非叶结点的索引
    char * word, *code ,codelen; //  分别是词的字面,霍夫曼编码,编码长度
};


// 训练文件、输出文件名称定义
char train_file[MAX_STRING], output_file[MAX_STRING];
// 词汇表输出文件和词汇表读入文件名称定义
char save_vocab_file[MAX_STRING], read_vocab_file[MAX_STRING];

// 声明词汇表结构体
struct vocab_word *vocab;
// binary 0则vectors.bin输出为二进制(默认),1则为文本形式
// cbow 1使用cbow框架,0使用skip-gram框架
// debug_mode 大于0,加载完毕后输出汇总信息,大于1,加载训练词汇的时候输出信息,训练过程中输出信息
// window 窗口大小,在cbow中表示了word vector的最大的sum范围,在skip-gram中表示了max space between words(w1,w2,p(w1 | w2))
// min_count 删除长尾词的词频标准
// num_threads 线程数
// min_reduce ReduceVocab删除词频小于这个值的词,因为哈希表总共可以装填的词汇数是有限的
int binary = 0, cbow = 0, debug_mode = 2, window = 5, min_count = 5, num_threads = 1, min_reduce = 1;
int *vocab_hash; // 词汇表的hash存储,下标是词的hash,内容是词在vocab中的位置,a[word_hash] = word index in vocab
// vocab_max_size 词汇表的最大长度,可以扩增,每次扩1000
// vocab_size 词汇表的现有长度,接近vocab_max_size的时候会扩容
// layer1_size 隐层的节点数
long long vocab_max_size = 1000, vocab_size = 0, layer1_size = 100;
// train_words 训练的单词总数(词频累加)
// word_count_actual 已经训练完的word个数
// file_size 训练文件大小,ftell得到
// classes 输出word clusters的类别数
long long train_words = 0, word_count_actual = 0, file_size = 0, classes = 0;
// alpha BP算法的学习速率,过程中自动调整
// starting_alpha 初始alpha值
// sample 亚采样概率的参数,亚采样的目的是以一定概率拒绝高频词,使得低频词有更多出镜率,默认为0,即不进行亚采样
real alpha = 0.025, starting_alpha, sample = 0;
// syn0 单词的向量输入 concatenate word vectors
// syn1 hs(hierarchical softmax)算法中隐层节点到霍夫曼编码树非叶结点的映射权重
// syn1neg ns(negative sampling)中隐层节点到分类问题的映射权重
// expTable 预先存储f函数结果,算法执行中查表
real *syn0, *syn1, *syn1neg, *expTable;
// start 算法运行的起始时间,会用于计算平均每秒钟处理多少词
clock_t start;

// hs 采用hs还是ns的标志位,默认采用hs
int hs = 1, negative = 0;
// table_size 静态采样表的规模
// table 采样表
const int table_size = 1e8;
int *table;

// 根据词频生成采样表
void InitUnigramTable() {
    int a, i;
    long long train_words_pow = 0;
    real d1, power = 0.75; // 概率与词频的power次方成正比
    table = (int *)malloc(table_size * sizeof(int));
    for (a = 0; a < vocab_size; a++) train_words_pow += pow(vocab[a].cn, power);
    i = 0;
    d1 = pow(vocab[i].cn, power) / (real)train_words_pow; // 第一个词出现的概率
    for (a = 0; a < table_size; a++) {
        table[a] = i;
        if (a / (real)table_size > d1) {
            i++;
            d1 += pow(vocab[i].cn, power) / (real)train_words_pow;
        }
        if (i >= vocab_size) i = vocab_size - 1; // 处理最后一段概率,防止越界
    }
}


//从文件file中每次读取一个单词
void ReadWord(char * word, FILE * file){

    int a = 0, ch;
    while (!feof(file)) {
        ch = fgetc(file);
        if (ch == 13)  continue; //读下一行 \r
        if ((ch == ' ') || (ch == '\t') || (ch == '\n') ) {
            if(a > 0){
                if (ch == '\n') ungetc(ch, file);// 把一个字符回退到输入流中
                break;
            }
            if (ch == '\n') {
                strcpy(word, (char *)"</s>");
                return;
            }else continue;
        }
        word[a] = ch;
        a++;
        if (a >= MAX_STRING - 1) a--;
    }
    word[a] = 0;
}

int GetWordHash(char * word) {
    unsigned long long a ,hash = 0;
    for (a = 0; a < strlen(word); a++) {
        hash = hash * 257 + word[a];
    }
    hash = hash % vocab_hash_size;
    return hash;
}

int AddWordToVocab(char * word){
    unsigned int hash, lengh = strlen(word) + 1;
    if (lengh > MAX_STRING) lengh = MAX_STRING;
    vocab[vocab_size].word = (char *) calloc(lengh, sizeof(char));
    strcpy(vocab[vocab_size].word, word);
    vocab[vocab_size].cn = 0;
    vocab_size++;
    if (vocab_size + 2 >= vocab_max_size ) {
        vocab_max_size += 1000;// 每次增加1000个词位
        vocab = (struct vocab_word *)realloc(vocab, vocab_max_size * sizeof(struct vocab_word));
    }
    hash = GetWordHash(word);
    while (vocab_hash[hash] != -1)  hash = (hash + 1) % vocab_hash_size; // 线性探索hash
    vocab_hash[hash] = vocab_size - 1;// 记录在词汇表中的存储位置
    return  vocab_size - 1;// 返回添加的单词在词汇表中的存储位置


}
//// 比较函数,词汇表需使用词频进行排序(qsort)
int  VocabCompare(const *a, const *b){
    return ((struct vocab_word *)b)->cn - ((struct vocab_word *)a)->cn;
}


//根据单词词频排序
void SortVocab() {
    int a, size;
    unsigned int hash;
    // 排序
    // 并且保证</s>在第一位
    qsort(&vocab[1], vocab_size - 1, sizeof(struct vocab_word), VocabCompare);//词汇表快排
    for (a = 0; a < vocab_hash_size; a++) vocab_hash[a] = -1;//词汇重排了,哈希记录的index也乱了,所有的hash记录清除,下面会重建
    size = vocab_size;
    train_words = 0;// 用于训练的词汇总数(词频累加)
    for (a = 0; a < size; a++) {
        // 删除特别低频的词
        if (vocab[a].cn < min_count) {
            vocab_size--;
            free(vocab[vocab_size].word);
        } else {
            //原来的hash失效需要重新计算
            hash=GetWordHash(vocab[a].word);
            while (vocab_hash[hash] != -1) hash = (hash + 1) % vocab_hash_size;
            vocab_hash[hash] = a;
            train_words += vocab[a].cn;
        }
    }
    vocab = (struct vocab_word *)realloc(vocab, (vocab_size + 1) * sizeof(struct vocab_word));
    // 给霍夫曼编码和路径的词汇表索引分配空间
    for (a = 0; a < vocab_size; a++) {
        vocab[a].code = (char *)calloc(MAX_CODE_LENGTH, sizeof(char));
        vocab[a].point = (int *)calloc(MAX_CODE_LENGTH, sizeof(int));
    }
}

// 读入词汇表文件到词汇表数据结构
void ReadVocab() {
    long long a, i = 0;
    char c;
    char word[MAX_STRING];
    FILE * fin = fopen(read_vocab_file, "rb");
    if (fin == NULL) {
        printf("Vocabulary file not found\n");
        exit(1);
    }
    for (a = 0; a < vocab_hash_size; a++)
        vocab_hash[a] = -1;
    vocab_size = 0;
    while (1) {
        ReadWord(word, fin);
        if (feof(fin)) break;
        fscanf(fin, "%lld%c", &vocab[a].cn, &c);
        i++;
    }
    SortVocab();
    if (debug_mode > 0) {
        printf("Vocab size: %lld\n", vocab_size);
        printf("Words in train file: %lld\n", train_words);
    }
    fin = fopen(train_file, "rb");
    if (fin == NULL) {
        printf("ERROR: training data file not found!\n");
        exit(1);
    }
    fseek(fin, 0, SEEK_END);
    file_size = ftell(fin);
    fclose(fin);
}

// 线性探索,开放定址法
int SearchVocab(char * word) {
    unsigned int hash = GetWordHash(word);
    while (1) {
        // 没有这个词
        if (vocab_hash[hash] == -1) return -1;
        if (!strcmp(word, vocab[vocab_hash[hash]].word)) return vocab_hash[hash];
        hash = (hash + 1) % vocab_hash_size;
    }
    return -1; // 额,这个代码永远运行不到......
}

void ReduceVocab() {
    // reduces the vocabulary by removing infrequent tokens.
    int a, b = 0;
    unsigned int hash;
    // 最后剩下b个词,词频均大于min_reduce
    for (a = 0; a < vocab_size; a++) {
        if (vocab[a].cn > min_reduce) {
            vocab[b].cn = vocab[a].cn;
            vocab[b].word = vocab[a].word;
            b++;
        } else {
            free (vocab[a].word);
        }
    }
    vocab_size = b;
    // 重新分配hash索引
    for (a = 0; a < vocab_hash_size; a++) vocab_hash[a] = -1;
    for (a = 0; a < vocab_size; a++) {
        hash = GetWordHash(vocab[a].word);
        while (vocab_hash[hash] != -1) hash = (hash + 1) % vocab_hash_size;
        vocab_hash[hash] = a;
    }
    fflush(stdout);
    min_reduce++;
}

// 装载训练文件到词汇表数据结构
void LearnVocabFromTrainFile() {
    char word[MAX_STRING];
    FILE * fin;
    long long a, i;
    for (a = 0; a < vocab_hash_size; a++) vocab_hash[a] = -1;
    fin = fopen(train_file, "rb");
    if (fin == NULL) {
        printf("ERROR: training data file not found!\n");
        exit(1);
    }
    vocab_size = 0;
    // 首先添加的是回车
    AddWordToVocab((char *)"</s>");
    while (1) {
        ReadWord(word, fin);
        if (feof(fin)) break;
        train_words++;
        if ((debug_mode > 1) && (train_words % 100000 == 0)) {
            printf("%lldK%c", train_words / 1000, 13);
            fflush(stdout);
        }
        i = SearchVocab(word);
        if (i == -1) {// 如果这个单词不存在,我们将其加入hash表
            a = AddWordToVocab(word);
            vocab[a].cn = 1;
        } else vocab[i].cn++;// 否则词频加一
        // 如果超出装填系数,将词汇表扩容
        if (vocab_size > vocab_hash_size * 0.7) ReduceVocab();
    }
    SortVocab();
    if (debug_mode > 0) {
        printf("Vocab size: %lld\n", vocab_size);
        printf("Words in train file: %lld\n", train_words);
    }
    file_size = ftell(fin);//文件大小
    fclose(fin);
}

// 输出单词和词频到文件
void SaveVocab() {
    long long i;
    FILE * fo = fopen(save_vocab_file, "wb");
    for (i = 0; i < vocab_size; i++) {
        fprintf(fo, "%s %lld\n", vocab[i].word, vocab[i].cn);
    }
    fclose(fo);
}

// 根据词频生成霍夫曼树
void CreateBinaryTree() {
    long long a, b, i;
    long long min1i, min2i;
    long long pos1, pos2;
    long long point[MAX_CODE_LENGTH];// 最长的编码值
    char code[MAX_CODE_LENGTH];
    long long *count = (long long *)calloc(vocab_size * 2 + 1, sizeof(long long));
    long long *binary = (long long *)calloc(vocab_size * 2 + 1, sizeof(long long));
    long long *parent_node = (long long *)calloc(vocab_size * 2 + 1, sizeof(long long));
    for (a = 0; a < vocab_size; a++) count[a] = vocab[a].cn;
    for (a = vocab_size; a < vocab_size * 2; a++) count[a] = 1e15;
    pos1 = vocab_size - 1;
    pos2 = vocab_size;
    for (a = 0; a < vocab_size - 1; a++) {
        // 每次寻找两个最小的点做合并,最小的点为0,词小的点为1
        if (pos1 >= 0) {
            if (count[pos1] < count[pos2]) {
                min1i = pos1;
                pos1--;
            } else {
                min1i = pos2;
                pos2++;
            }
        } else {
            min1i = pos2;
            pos2++;
        }
        if (pos1 >= 0) {
            if (count[pos1] < count[pos2]) {
                min2i = pos1;
                pos1--;
            } else {
                min2i = pos2;
                pos2++;
            }
        } else {
            min2i = pos2;
            pos2++;
        }
        count[vocab_size + a] = count[min1i] + count[min2i];
        parent_node[min1i] = vocab_size + a;
        parent_node[min2i] = vocab_size + a;
        binary[min2i] = 1;
    }
    // 顺着父子关系找回编码
    for (a = 0; a < vocab_size; a++) {
        b = a;
        i = 0;
        while (1) {
            code[i] = binary[b];
            point[i] = b;

            i++;
            b = parent_node[b];
            if (b == vocab_size * 2 - 2) break;
        }
        vocab[a].codelen = i;
        vocab[a].point[0] = vocab_size - 2; // 逆序,把第一个赋值为root(即2*vocab_size - 2 - vocab_size)
        for (b = 0; b < i; b++) {
            vocab[a].code[i - b - 1] = code[b];// 编码逆序,没有根节点,左子树0,右子树1
            vocab[a].point[i - b] = point[b] - vocab_size;
        }
    }
    free(count);
    free(binary);
    free(parent_node);
}


int ReadWordIndex(FILE * fin) {
    char word[MAX_STRING];
    ReadWord(word, fin);
    if (feof(fin)) return -1;
    return SearchVocab(word);
}

// 网络结构初始化
void InitNet() {
    // intialize the neural network structure
    long long a, b;
    // posix_memalign() 成功时会返回size字节的动态内存,并且这块内存的地址是alignment(这里是128)的倍数
    // syn0 存储的是word vectors
    a = posix_memlign((void **)&syn0, 128, (long long)vocab_size * layer1_size * sizeof(real));
    if (syn0 == NULL) {
        printf("Memory allocation failed\n"); exit(1);
    }
    // Hierarchical Softmax
    if (hs) {
        // hs中,用syn1
        a = posix_memlign((void **)&syn1, 128, (long long)vocab_size * layer1_size * sizeof(real));
        if (syn1 == NULL) {printf("Memory allocation failed\n"); exit(1);}
        for (b = 0; b < layer1_size; b++) for (a = 0; a < vocab_size; a++)
            syn1[a * layer1_size + b] = 0;
    }
    // Negative Sampling
    if (negative > 0) {
        // ns中,用syn1neg
        a = posix_memlign((void **)&syn1neg, 128, (long long)vocab_size * layer1_size * sizeof(real));
        if (syn1neg == NULL) {printf("Memory allocaiton failed\n"); exit(1);}
        for (b = 0; b < layer1_size; b++) for (a = 0; a < vocab_size; a++)
            syn1neg[a * layer1_size + b] = 0;
    }
    for (b = 0; b < layer1_size; b++) for (a = 0; a < vocab_size; a++)
        syn0[a * layer1_size + b] = (rand() / (real)RAND_MAX - 0.5) / layer1_size;//随机初始化word vectors
    CreateBinaryTree();// 创建霍夫曼树
}

// learning: hs (hierarchical softmax) v.s. negative sampling
// model: cbow v.s. skip gram
void *TrainModelThread(void *id) {
    // word 向sen中添加单词用,句子完成后表示句子中的当前单词
    // last_word 上一个单词,辅助扫描窗口
    // sentence_length 当前句子的长度(单词数)
    // sentence_position 当前单词在当前句子中的index
    long long a, b, d, word, last_word, sentence_length = 0, sentence_position = 0;
    // word_count 已训练语料总长度
    // last_word_count 保存值,以便在新训练语料长度超过某个值时输出信息
    // sen 单词数组,表示句子
    long long word_count = 0, last_word_count = 0, sen[MAX_SENTENCE_LENGTH + 1];
    // l1 ns中表示word在concatenated word vectors中的起始位置,之后layer1_size是对应的word vector,因为把矩阵拉成长向量了
    // l2 cbow或ns中权重向量的起始位置,之后layer1_size是对应的syn1或syn1neg,因为把矩阵拉成长向量了
    // c 循环中的计数作用
    // target ns中当前的sample
    // label ns中当前sample的label

    long long l1, l2, c, target, label;
    // id 线程创建的时候传入,辅助随机数生成
    unsigned long long next_random = (long long) id;
    // f e^x / (1/e^x),fs中指当前编码为是0(父亲的左子节点为0,右为1)的概率,ns中指label是1的概率
    // g 误差(f与真实值的偏离)与学习速率的乘积
    real f, g; // function and gradient
    clock_t now;
    // 隐层节点
    real * neu1 = (real *)calloc(layer1_size, sizeof(real));
    // 误差累计项,其实对应的是Gneu1
    real * neu1e = (real *)calloc(layer1_size, sizeof(real));
    // 将文件内容分配给各个线程
    FILE * fi = fopen(train_file, "rb");
    fseek(fi, file_size / (long long)num_threads * (long long)id, SEEK_SET);

    while (1) {
        if (word_count - last_word_count > 10000) {
            word_count_actual += word_count - last_word_count;
            last_word_count = word_count;
            if (debug_mode > 1) {
                now = clock();
                printf("%cAlpah: %f Progress: %.2f%% Words/thread/sec: %.2fk ", 13, alpha,
                       word_count_actual / (real)(train_words + 1) * 100,
                       word_count_actual / ((real)(now - start + 1) / (real)CLOCKS_PER_SEC * 1000));
                fflush(stdout);
            }
            alpha = starting_alpha * (1 - word_count_actual / (real)(train_words + 1)); // 自动调整学习速率
            if (alpha < starting_alpha * 0.0001) alpha = starting_alpha * 0.0001;// 学习速率有下限
        }
        if (sentence_length == 0) {// 如果当前句子长度为0
            while(1) {
                word = ReadWordIndex(fi);
                if (feof(fi)) break;// 读到文件末尾
                if (word == -1) continue;// 没有这个单词
                word_count++;// 单词计数增加
                if (word == 0) break;// 是个回车
                // 这里的亚采样是指 Sub-Sampling,Mikolov 在论文指出这种亚采样能够带来 2 到 10 倍的性能提升,并能够提升低频词的表示精度。
                // 低频词被丢弃概率低,高频词被丢弃概率高
                if (sample > 0) {
                    real ran = (sqrt(vocab[word].cn / (sample * train_words)) + 1) * (sample * train_words) / vocab[word].cn;
                    next_random = next_random * (unsigned long long)25214903917 + 11;
                    if (ran < (next_random & 0xFFFF) / (real)65536) continue;
                }
                sen[sentence_length] = word;
                sentence_length++;
                if (sentence_length >= MAX_SENTENCE_LENGTH) break;
            }
            sentence_position = 0;// 当前单词在当前句中的index,起始值为0
        }
        // 照应while中的break,如果读到末尾,退出
        if (feof(fi)) break;
        // 已经做到了一个thread应尽的工作量,就退出
        if (word_count > train_words / num_threads) break;
        // 取句子中的第一个单词,开始运行BP算法
        word = sen[sentence_position];
        if (word == -1) continue;
        // 隐层节点值和隐层节点误差累计项清零
        for (c = 0; c < layer1_size; c++) neu1[c] = 0;
        for (c = 0; c < layer1_size; c++) neu1e[c] = 0;
        next_random = next_random * (unsigned long long)25214903917 + 11;
        // b是个随机数,0到window-1,指定了本次算法操作实际的窗口大小
        b = next_random % window;

        if (cbow) { // train the cbow architecture - HS or NS
            // IN -> HIDDEN
            // 将窗口内的word vectors累加到隐层节点上
            for (a = b; a < window * 2 + 1 - b; a++) if (a != window) {
                c = sentence_position - window + a;
                if (c < 0) continue;
                if (c >= sentence_length) continue;
                last_word = sen[c];
                if (last_word == -1) continue;// 这个单词没有
                for (c = 0; c < layer1_size; c++)
                    neu1[c] += syn0[c + last_word * layer1_size];
            }
            // HIERARCHICAL SOFTMAX
            if (hs) for (d = 0; d < vocab[word].codelen; d++) {
                // 这里的codelen其实是少一个的,所以不会触及point里面最后一个负数
                f = 0;
                l2 = vocab[word].point[d] * layer1_size;
                // propagate hidden -> output
                // 准备计算f
                for (c = 0; c < layer1_size; c++) f += neu1[c] * syn1[c + l2];
                // 不在expTable内的舍弃掉,比较暴力
                if (f <= -MAX_EXP) continue;
                else if (f >= MAX_EXP) continue;
                // 从expTable中查找,快速计算
                else f = expTable[(int)((f + MAX_EXP) * (EXP_TABLE_SIZE / MAX_EXP / 2))];
                // g is the gradient multiplied by the learning rate
                g = (1 - vocab[word].code[d] -f) * alpha;
                // propogate errors output -> hidden
                 // 记录累积误差项
                for (c = 0; c < layer1_size; c++) neu1e[c] += g * syn1[c + l2];
                // learning weights hidden -> output
                // 更新隐层到霍夫曼树非叶节点的权重
                for (c = 0; c < layer1_size; c++) syn1[c + l2] += g * neu1[c];
            }
            // NEGATIVE SAMPLING
            if (negative > 0) for (d = 0; d < negative + 1; d++) {
                if (d == 0) { // 当前词的分类器应当输出1
                    target = word;
                    label = 1;
                } else { // 采样使得与target不同,不然continue,label为0,也即最多采样negative个negative sample
                    next_random = next_random * (unsigned long long)25214903917 + 11;
                    target = table[(next_random >> 16) % table_size];
                    if (target == 0) target = next_random % (vocab_size - 1) + 1;
                    if (target == word) continue;
                    label = 0;
                }
                l2 = target * layer1_size;
                f = 0;
                for (c = 0; c < layer1_size; c++) f += neu1[c] * syn1neg[c + l2];
                // 这里直接上0、1,没有考虑计算精度问题……
                if (f > MAX_EXP) g = (label - 1) * alpha;
                else if (f < -MAX_EXP) g = (label - 0) * alpha;
                else g = (label - expTable[(int)((f + MAX_EXP) * (EXP_TABLE_SIZE / MAX_EXP / 2))]) * alpha;
                for (c = 0; c < layer1_size; c++) neu1e[c] += g * syn1neg[c + l2];
                for (c = 0; c < layer1_size; c++) syn1neg[c + l2] += g * neu1[c];
            }
            // hidden -> in
            // 根据隐层节点累积误差项,更新word vectors
            for (a = b; a < window * 2 + 1 - b; a++) if (a != window) {
                c = sentence_position - window + a;
                if (c < 0) continue;
                if (c >= sentence_length) continue;
                last_word = sen[c];
                if (last_word == -1) continue;
                for (c = 0; c < layer1_size; c++) syn0[c + last_word * layer1_size] += neu1e[c];
            }
        } else {  //train skip-gram
            for (a = b; a < window * 2 + 1 - b; a++) if (a != window) { // 预测非中心的单词(邻域内的单词)
                c = sentence_position - window + a;
                if (c < 0) continue;
                if (c >= sentence_length) continue;
                last_word = sen[c];
                if (last_word == -1) continue;
                l1 = last_word * layer1_size;
                // 累计误差项清零
                for (c = 0; c < layer1_size; c++) neu1e[c] = 0;

                // HIERARCHICAL SOFTMAX
                if (hs) for (d = 0; d < vocab[word].codelen; d++) {
                    f = 0;
                    l2 = vocab[word].point[d] * layer1_size;
                    // Propagate hidden -> output
                    // 待预测单词的 word vecotr 和 隐层-霍夫曼树非叶节点权重 的内积
                    for (c = 0; c < layer1_size; c++) f += syn0[c + l1] * syn1[c + l2];
                    // 同cbow中hs的讨论
                    // if (f <= -MAX_EXP) continue;
                    // else if (f >= MAX_EXP) continue;
                    if (f <= -MAX_EXP) f = 0;
                    else if (f >= MAX_EXP) f = 1;
                    // 以下内容同之前的cbow
                    else f = expTable[(int)((f + MAX_EXP) * (EXP_TABLE_SIZE / MAX_EXP / 2))];
                    // 'g' is the gradient multiplied by the learning rate
                    g = (1 - vocab[word].code[d] - f) * alpha; // 这里的code[d]其实是下一层的,code错位了,point和code是错位的!
                    // Propagate errors output -> hidden
                    for (c = 0; c < layer1_size; c++) neu1e[c] += g * syn1[c + l2];
                    // Learn weights hidden -> output
                    for (c = 0; c < layer1_size; c++) syn1[c + l2] += g * syn0[c + l1];
                }
                // NEGATIVE SAMPLING
                if (negative > 0) for (d = 0; d < negative + 1; d++) {
                    if (d == 0) {
                        target = word;
                        label = 1;
                    } else {
                        next_random = next_random * (unsigned long long)25214903917 + 11;
                        target = table[(next_random >> 16) % table_size];
                        if (target == 0) target = next_random % (vocab_size - 1) + 1;
                        if (target == word) continue;
                        label = 0;
                    }
                    l2 = target * layer1_size;
                    f = 0;
                    for (c = 0; c < layer1_size; c++) f += syn0[c + l1] * syn1neg[c + l2];
                    // 以下内容同之前的cbow
                    if (f > MAX_EXP) g = (label - 1) * alpha;
                    else if (f < -MAX_EXP) g = (label - 0) * alpha;
                    else g = (label - expTable[(int)((f + MAX_EXP) * (EXP_TABLE_SIZE / MAX_EXP / 2))]) * alpha;
                    for (c = 0; c < layer1_size; c++) neu1e[c] += g * syn1neg[c + l2];
                    for (c = 0; c < layer1_size; c++) syn1neg[c + l2] += g * syn0[c + l1];
                }
                // Learn weights input -> hidden
                for (c = 0; c < layer1_size; c++) syn0[c + l1] += neu1e[c];
            }
        }
        sentence_position++;
        if (sentence_position >= sentence_length) {
            sentence_length = 0;
            continue;
        }
    }
    fclose(fi);
    free(neu1);
    free(neu1e);
    pthread_exit(NULL);
}


void TrainModel() {
    long a, b, c, d;
    FILE *fo;
    // 创建多线程
    pthread_t *pt = (pthread_t *)malloc(num_threads * sizeof(pthread_t));
    printf("Starting training using file %s\n", train_file);
    starting_alpha = alpha;
    // 优先从词汇表文件中加载,否则从训练文件中加载
    if (read_vocab_file[0] != 0) ReadVocab(); else LearnVocabFromTrainFile();
    // 输出词汇表文件,词+词频
    if (save_vocab_file[0] != 0) SaveVocab();
    if (output_file[0] == 0) return;
    InitNet(); // 网络结构初始化
    if (negative > 0) InitUnigramTable(); // 根据词频生成采样映射
    start = clock(); // 开始计时
    for (a = 0; a < num_threads; a++) pthread_create(&pt[a], NULL, TrainModelThread, (void *)a);
    for (a = 0; a < num_threads; a++) pthread_join(pt[a], NULL);

    // 训练结束,准备输出
    fo = fopen(output_file, "wb");

    if (classes == 0) { // 保存 word vectors
        // Save the word vectors
        fprintf(fo, "%lld %lld\n", vocab_size, layer1_size); // 词汇量,vector维数
        for (a = 0; a < vocab_size; a++) {
            fprintf(fo, "%s ", vocab[a].word);
            if (binary) for (b = 0; b < layer1_size; b++) fwrite(&syn0[a * layer1_size + b], sizeof(real), 1, fo);
            else for (b = 0; b < layer1_size; b++) fprintf(fo, "%lf ", syn0[a * layer1_size + b]);
            fprintf(fo, "\n");
        }
    } else {
        // Run K-means on the word vectors
        // 运行K-means算法
        int clcn = classes, iter = 10, closeid;
        int *centcn = (int *)malloc(classes * sizeof(int));
        int *cl = (int *)calloc(vocab_size, sizeof(int));
        real closev, x;
        real *cent = (real *)calloc(classes * layer1_size, sizeof(real));
        for (a = 0; a < vocab_size; a++) cl[a] = a % clcn;
        for (a = 0; a < iter; a++) {
            for (b = 0; b < clcn * layer1_size; b++) cent[b] = 0;
            for (b = 0; b < clcn; b++) centcn[b] = 1;
            for (c = 0; c < vocab_size; c++) {
                for (d = 0; d < layer1_size; d++) cent[layer1_size * cl[c] + d] += syn0[c * layer1_size + d];
                centcn[cl[c]]++;
            }
            for (b = 0; b < clcn; b++) {
                closev = 0;
                for (c = 0; c < layer1_size; c++) {
                    cent[layer1_size * b + c] /= centcn[b];
                    closev += cent[layer1_size * b + c] * cent[layer1_size * b + c];
                }
                closev = sqrt(closev);
                for (c = 0; c < layer1_size; c++) cent[layer1_size * b + c] /= closev;
            }
            for (c = 0; c < vocab_size; c++) {
                closev = -10;
                closeid = 0;
                for (d = 0; d < clcn; d++) {
                    x = 0;
                    for (b = 0; b < layer1_size; b++) x += cent[layer1_size * d + b] * syn0[c * layer1_size + b];
                    if (x > closev) {
                        closev = x;
                        closeid = d;
                    }
                }
                cl[c] = closeid;
            }
        }
        // Save the K-means classes
        for (a = 0; a < vocab_size; a++) fprintf(fo, "%s %d\n", vocab[a].word, cl[a]);
        free(centcn);
        free(cent);
        free(cl);
    }
    fclose(fo);
}

int ArgPos(char *str, int argc, char **argv) {
    int a;
    for (a = 1; a < argc; a++) if (!strcmp(str, argv[a])) {
        if (a == argc - 1) {
            printf("Argument missing for %s\n", str);
            exit(1);
        }
        return a;
    }
    return -1;
}

int main(int argc, char **argv) {
    int i;
    if (argc == 1) {
        printf("WORD VECTOR estimation toolkit v 0.1b\n\n");
        printf("Options:\n");
        printf("Parameters for training:\n");
        printf("\t-train <file>\n"); // 指定训练文件
        printf("\t\tUse text data from <file> to train the model\n");
        printf("\t-output <file>\n"); // 指定输出文件,以存储word vectors,或者单词类
        printf("\t\tUse <file> to save the resulting word vectors / word clusters\n");
        printf("\t-size <int>\n"); // word vector的维数,对应 layer1_size,默认是100
        printf("\t\tSet size of word vectors; default is 100\n");
        // 窗口大小,在cbow中表示了word vector的最大的叠加范围,在skip-gram中表示了max space between words(w1,w2,p(w1 | w2))
        printf("\t-window <int>\n");
        printf("\t\tSet max skip length between words; default is 5\n");
        printf("\t-sample <float>\n"); // 亚采样拒绝概率的参数
        printf("\t\tSet threshold for occurrence of words. Those that appear with higher frequency");
        printf(" in the training data will be randomly down-sampled; default is 0 (off), useful value is 1e-5\n");
        printf("\t-hs <int>\n"); // 使用hs求解,默认为1
        printf("\t\tUse Hierarchical Softmax; default is 1 (0 = not used)\n");
        printf("\t-negative <int>\n"); // 使用ns的时候采样的样本数
        printf("\t\tNumber of negative examples; default is 0, common values are 5 - 10 (0 = not used)\n");
        printf("\t-threads <int>\n"); // 指定线程数
        printf("\t\tUse <int> threads (default 1)\n");
        printf("\t-min-count <int>\n"); // 长尾词的词频阈值
        printf("\t\tThis will discard words that appear less than <int> times; default is 5\n");
        printf("\t-alpha <float>\n"); // 初始的学习速率,默认为0.025
        printf("\t\tSet the starting learning rate; default is 0.025\n");
        printf("\t-classes <int>\n"); // 输出单词类别数,默认为0,也即不输出单词类
        printf("\t\tOutput word classes rather than word vectors; default number of classes is 0 (vectors are written)\n");
        printf("\t-debug <int>\n"); // 调试等级,默认为2
        printf("\t\tSet the debug mode (default = 2 = more info during training)\n");
        printf("\t-binary <int>\n"); // 是否将结果输出为二进制文件,默认为0,即不输出为二进制
        printf("\t\tSave the resulting vectors in binary moded; default is 0 (off)\n");
        printf("\t-save-vocab <file>\n"); // 词汇表存储文件
        printf("\t\tThe vocabulary will be saved to <file>\n");
        printf("\t-read-vocab <file>\n"); // 词汇表加载文件,则可以不指定trainfile
        printf("\t\tThe vocabulary will be read from <file>, not constructed from the training data\n");
        printf("\t-cbow <int>\n"); // 使用cbow框架
        printf("\t\tUse the continuous bag of words model; default is 0 (skip-gram model)\n");
        printf("\nExamples:\n"); // 使用示例
        printf("./word2vec -train data.txt -output vec.txt -debug 2 -size 200 -window 5 -sample 1e-4 -negative 5 -hs 0 -binary 0 -cbow 1\n\n");
        return 0;
    }
    // 文件名均空
    output_file[0] = 0;
    save_vocab_file[0] = 0;
    read_vocab_file[0] = 0;
    // 参数与变量的对应关系
    if ((i = ArgPos((char *)"-size", argc, argv)) > 0) layer1_size = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-train", argc, argv)) > 0) strcpy(train_file, argv[i + 1]);
    if ((i = ArgPos((char *)"-save-vocab", argc, argv)) > 0) strcpy(save_vocab_file, argv[i + 1]);
    if ((i = ArgPos((char *)"-read-vocab", argc, argv)) > 0) strcpy(read_vocab_file, argv[i + 1]);
    if ((i = ArgPos((char *)"-debug", argc, argv)) > 0) debug_mode = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-binary", argc, argv)) > 0) binary = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-cbow", argc, argv)) > 0) cbow = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-alpha", argc, argv)) > 0) alpha = atof(argv[i + 1]);
    if ((i = ArgPos((char *)"-output", argc, argv)) > 0) strcpy(output_file, argv[i + 1]);
    if ((i = ArgPos((char *)"-window", argc, argv)) > 0) window = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-sample", argc, argv)) > 0) sample = atof(argv[i + 1]);
    if ((i = ArgPos((char *)"-hs", argc, argv)) > 0) hs = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-negative", argc, argv)) > 0) negative = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-threads", argc, argv)) > 0) num_threads = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-min-count", argc, argv)) > 0) min_count = atoi(argv[i + 1]);
    if ((i = ArgPos((char *)"-classes", argc, argv)) > 0) classes = atoi(argv[i + 1]);

    vocab = (struct vocab_word *)calloc(vocab_max_size, sizeof(struct vocab_word));
    vocab_hash = (int *)calloc(vocab_hash_size, sizeof(int));
    expTable = (real *)malloc((EXP_TABLE_SIZE + 1) * sizeof(real));
    // 产生e^-6 到 e^6 之间的f值
    for (i = 0; i < EXP_TABLE_SIZE; i++) {
        expTable[i] = exp((i / (real)EXP_TABLE_SIZE * 2 - 1) * MAX_EXP); // Precompute the exp() table
        expTable[i] = expTable[i] / (expTable[i] + 1);                   // Precompute f(x) = x / (x + 1)
    }
    TrainModel();
    return 0;
}
点击查看评论