课程链接：点这里
数据集：课程评论区有课件链接，里面有数据集
刘洪普老师个人网站：点这里

1. GPU版本的程序：

1.1 环境要求

python3.8.3
torch==1.7.1+cu101
cuda 10.1
显卡: Tesla V100

1.2 程序

import math
import time
import torch
import torch.nn as nn
import torch.optim as optim
from torch.utils.data import Dataset
from torch.utils.data import DataLoader
from torch.nn.utils.rnn import pack_padded_sequence
import gzip
import csv
import matplotlib.pyplot as plt
import numpy as np


# 读取数据集以及获取数据集信息
class NameDataset(Dataset):
    def __init__(self, is_train_set=True):
        filename = '/home/Datasets/names_train.csv.gz' if is_train_set else '/home/Datasets/names_test.csv.gz'
        with gzip.open(filename, 'rt') as f:
            reader = csv.reader(f)
            rows = list(reader)
        # 取出来名字那一列
        self.names = [row[0] for row in rows]
        self.len = len(self.names)
        self.countries = [row[1] for row in rows]
        # set去除重复元素
        self.country_list = list(sorted(set(self.countries)))
        self.country_dict = self.getCountryDict()
        self.country_num = len(self.country_list)

    # 根据索引获取元素
    def __getitem__(self, index):
        return self.names[index], self.country_dict[self.countries[index]]

    def __len__(self):
        return self.len

    """ 作用：生成无重复的国家字典
    # 示例：
        countries = ['China', 'America', 'Jap', 'China', 'England', 'Jap', 'Koera', 'Russian', 'America', 'India', 'India', 'China', 'Germ', 'Franch', 'Germ', 'Austra', 'Franch']
        country_list = list(sorted(set(countries)))
        print(country_list)


        def getCountryDict():
            country_dict = dict()
            for idx, country_name in enumerate(country_list, 0):
                country_dict[country_name] = idx
            return country_dict

        country_dict = getCountryDict()
        print(country_dict)

    # 输出：
    >> ['America', 'Austra', 'China', 'England', 'Franch', 'Germ', 'India', 'Jap', 'Koera', 'Russian']
    >> {'America': 0, 'Austra': 1, 'China': 2, 'England': 3, 'Franch': 4, 'Germ': 5, 'India': 6, 'Jap': 7, 'Koera': 8, 'Russian': 9}
     """
    def getCountryDict(self):
        country_dict = dict()
        for idx, country_name in enumerate(self.country_list, 0):
            country_dict[country_name] = idx
        return country_dict

    # 根据索引返回国家的字符串
    def idx2country(self, index):
        return self.country_list[index]

    def getCountriesNum(self):
        return self.country_num


class RNNClassifier(nn.Module):

    # __init__参数是底下所有的搭建网络所需要用到的参数
    # bidirection:双向GRU
    def __init__(self, input_size, hidden_size, output_size, n_layers=1, bidirectional=True):
        super(RNNClassifier, self).__init__()
        self.hidden_size = hidden_size
        self.n_layers = n_layers
        self.n_directions = 2 if bidirectional else 1

        self.embedding = nn.Embedding(input_size, hidden_size)
        self.gru = nn.GRU(hidden_size, hidden_size, n_layers, bidirectional=bidirectional)
        self.fc = nn.Linear(hidden_size * self.n_directions, output_size)

    # 初始化输入的数据
    def _init_hidden(self, batch_size):
        hidden = torch.zeros(self.n_layers * self.n_directions, batch_size, self.hidden_size)

        return create_tensor(hidden)

    def forward(self, input, seq_lengths):

        # 转置矩阵 将Batch*Seq转换成Seq*Batch  适合Embedding需要
        input = input.t()
        # input_data = input_data.transpose(0,1)

        batch_size = input.size(1)

        hidden = self._init_hidden(batch_size)
        # 形状：(seqLen, batch_size, hidden_size)
        embedding = self.embedding(input)  

        # pack_padded_sequence作用：去掉全零的序列，节省空间，
        gru_input = pack_padded_sequence(embedding, seq_lengths.cpu())

        output, hidden = self.gru(gru_input, hidden)
        if self.n_directions ==2:
            hidden_cat = torch.cat([hidden[-1], hidden[-2]], dim=-1)
        else:
            hidden_cat = hidden[-1]
        fc_output = self.fc(hidden_cat)

        return fc_output


def names2list(name):

    # ord()返回对应ASCLL表示
    arr = [ord(c) for c in name]
    return arr, len(arr)


def create_tensor(tensor):
    if USE_GPU:
        device = torch.device('cuda:0')
        tensor = tensor.to(device)
    return tensor


def time_since(since):
    s = time.time() - since
    m = math.floor(s / 60)
    s -= m * 60
    return '%dm %ds' %(m, s)

def make_tensor(names, countries):
    sequence_and_lengths = [names2list(name) for name in names]
    name_sequences = [sl[0] for sl in sequence_and_lengths]
    seq_lengths = torch.LongTensor([sl[1] for sl in sequence_and_lengths])
    countries = countries.long()

    # 先设置一个全零数组 然后把ASCLL数据替代过去
    seq_tensor = torch.zeros(len(name_sequences), seq_lengths.max()).long()
    # 遍历名字、序列长度组成的元组，从0开始 
    for idx, (seq, seq_len) in enumerate(zip(name_sequences, seq_lengths), 0):
        seq_tensor[idx, :seq_len] = torch.LongTensor(seq)
    """ 
    # 示例：
        import torch
        seq_lengths = torch.LongTensor([10, 9 , 10, 7, 8])
        seq_lengths, perm_idx = seq_lengths.sort(dim=0, descending=True)
        seq_lengths, perm_idx
    # 输出
    >> (tensor([10, 10,  9,  8,  7]), tensor([0, 2, 1, 4, 3]))
     """
    seq_lengths, perm_idx = seq_lengths.sort(dim=0, descending=True)
    # 将ASCLL表按照所需的顺序排列好
    seq_tensor = seq_tensor[perm_idx]
    # 将城市的名字按照所需的顺序排列好
    countries = countries[perm_idx]

    return create_tensor(seq_tensor), create_tensor(seq_lengths), create_tensor(countries)


# 训练模型
def trainModel():
    total_loss = 0
    
    """ 
    训练集13374个名字，BATCH_SIZE=256 = 53
    i的范围[1, 53]
     """
    for i, (names, countries) in enumerate(trainloader, 1):
        inputs, seq_lengths, target = make_tensor(names, countries)
        output = classifier(inputs, seq_lengths)
        loss = criterion(output, target)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        total_loss += loss.item()
        if i % 10 == 0:
            print(f'[{time_since(start)}] Epoch {epoch}', end=' ')
            print(f'[{i * len(inputs)}/{len(trainset)}]', end=' ')
            print(f'loss = {total_loss / (i * len(inputs))}')
    return total_loss


def testModel():
    correct = 0
    total = len(testset)
    print('Evaluating trained model...')
    with torch.no_grad():
        for i, (names, countries) in enumerate(testloader, 1):
            inputs, seq_lengths, target = make_tensor(names, countries)
            output = classifier(inputs, seq_lengths)
            pred = output.max(dim=1, keepdim=True)[1]
            correct += pred.eq(target.view_as(pred)).sum().item()

        percent = '%.2f' % (100 * correct / total)
        print(f'Test set: Accuracy {correct} / {total} {percent}%')
    return correct / total


if __name__ == '__main__':
    HIDDEN_SIZE = 100
    BATCH_SIZE = 256  # 每个batch_size256个名字在
    N_LAYERS = 2
    N_EPOCHS = 100  # 训练100轮
    N_CHARS = 128  # 字典长度128
    USE_GPU = True
    # USE_GPU = False

    trainset = NameDataset(is_train_set=True)
    trainloader = DataLoader(trainset, batch_size=BATCH_SIZE, shuffle=True)

    testset = NameDataset(is_train_set=False)
    testloader = DataLoader(testset, batch_size=BATCH_SIZE, shuffle=False)

    N_COUNTRY = trainset.getCountriesNum()

    classifier = RNNClassifier(N_CHARS, HIDDEN_SIZE, N_COUNTRY, N_LAYERS)
    if USE_GPU:
        device = torch.device('cuda:0')
        classifier.to(device)

    criterion = nn.CrossEntropyLoss()
    optimizer = optim.Adam(classifier.parameters(), lr=0.001)

    start = time.time()
    print('Training for %d epochs...' %N_EPOCHS)

    acc_list = []
    # 每训练一个Epoch就验证一下准确率
    for epoch in range(1, N_EPOCHS + 1):
        trainModel()
        acc = testModel()
        acc_list.append(acc)
    
    # 绘图，展示训练的结果，发现训练到N_EPOCHS=10的时候就效果不再显著
    epoch = np.arange(1, len(acc_list) + 1, 1)
    acc_list = np.array(acc_list)
    fig = plt.figure()
    ax = fig.add_subplot()
    ax.plot(epoch, acc_list, label='Accuracy')
    ax.set_title('Change of Accuracy')
    ax.set_xlabel('Epoch')
    ax.set_ylabel('Acc')
    ax.legend()
    ax.grid()
    plt.show()

gru_input = pack_padded_sequence(embedding, seq_lengths.cpu())
.cpu() 课程中没有，但是不加无法在GPU运行

1.3 运行结果

.....
[3m 13s] Epoch 100[2560/13374]loss = 0.00012714920594589785
[3m 14s] Epoch 100[5120/13374]loss = 0.00013297978839545976
[3m 14s] Epoch 100[7680/13374]loss = 0.00013881954461491356
[3m 14s] Epoch 100[10240/13374]loss = 0.00014577317033399596
[3m 15s] Epoch 100[12800/13374]loss = 0.00015000603394582868
Evaluating trained model...
Test set: Accuracy 5591 / 6700 83.45%

2. CPU版本的程序：

2.1 修改

USE_GPU = False

.....
[22m 10s] Epoch 100[2560/13374]loss = 0.00010133394935110118
[22m 13s] Epoch 100[5120/13374]loss = 0.0001161873473392916
[22m 15s] Epoch 100[7680/13374]loss = 0.000131447339784548
[22m 17s] Epoch 100[10240/13374]loss = 0.00014009950637046132
[22m 20s] Epoch 100[12800/13374]loss = 0.00014168857444019523
Evaluating trained model...
Test set: Accuracy 5594 / 6700 83.49%

可以看出来GPU加速效果还是很明显的，大概快了7倍