本篇不涉及模型原理，只是分享下代码。想要了解模型原理的可以去看网上很多大牛的博客。

目前代码实现了CNN和LSTM两个网络，整个代码分为四部分：

• Config：项目中涉及的参数；

• CNN：卷积神经网络结构；

• LSTM：长短期记忆网络结构；

• TrainProcess：

  模型训练及评估，参数model控制训练何种模型（CNN or LSTM）。

完整代码

—Talk is cheap, show me the code.

# -*- coding: utf-8 -*-

# @author: Awesome_Tang
# @date: 2019-04-05
# @version: python3.7

import torch
from torchvision import datasets, transforms
import torch.nn as nn
import torch.optim as optim
from torch.autograd import Variable
from datetime import datetime


class Config:
    batch_size = 64
    epoch = 10
    alpha = 1e-3

    print_per_step = 100  # 控制输出


class CNN(nn.Module):

    def __init__(self):
        super(CNN, self).__init__()
        """
        Conv2d参数：
        第一位：input channels  输入通道数
        第二位：output channels 输出通道数
        第三位：kernel size 卷积核尺寸
        第四位：stride 步长，默认为1
        第五位：padding size 默认为0，不补
        """
        self.conv1 = nn.Sequential(
            nn.Conv2d(1, 32, 3, 1, 2),
            nn.ReLU(),
            nn.MaxPool2d(2, 2)
        )

        self.conv2 = nn.Sequential(
            nn.Conv2d(32, 64, 5),
            nn.ReLU(),
            nn.MaxPool2d(2, 2)
        )

        self.fc1 = nn.Sequential(
            nn.Linear(64 * 5 * 5, 128),
            nn.BatchNorm1d(128),
            nn.ReLU()
        )

        self.fc2 = nn.Sequential(
            nn.Linear(128, 64),
            nn.BatchNorm1d(64),  # 加快收敛速度的方法（注：批标准化一般放在全连接层后面，激活函数层的前面）
            nn.ReLU()
        )

        self.fc3 = nn.Linear(64, 10)

    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        x = x.view(x.size()[0], -1)
        x = self.fc1(x)
        x = self.fc2(x)
        x = self.fc3(x)
        return x


class LSTM(nn.Module):
    def __init__(self):
        super(LSTM, self).__init__()

        self.lstm = nn.LSTM(
            input_size=28,
            hidden_size=64,
            num_layers=1,
            batch_first=True,
        )

        self.output = nn.Linear(64, 10)

    def forward(self, x):
        r_out, (_, _) = self.lstm(x, None)

        out = self.output(r_out[:, -1, :])
        return out


class TrainProcess:

    def __init__(self, model="CNN"):
        self.train, self.test = self.load_data()
        self.model = model
        if self.model == "CNN":
            self.net = CNN()
        elif self.model == "LSTM":
            self.net = LSTM()
        else:
            raise ValueError('"CNN" or "LSTM" is expected, but received "%s".' % model)
        self.criterion = nn.CrossEntropyLoss()  # 定义损失函数
        self.optimizer = optim.Adam(self.net.parameters(), lr=Config.alpha)

    @staticmethod
    def load_data():
        print("Loading Data......")
        """加载MNIST数据集，本地数据不存在会自动下载"""
        train_data = datasets.MNIST(root='./data/',
                                    train=True,
                                    transform=transforms.ToTensor(),
                                    download=True)

        test_data = datasets.MNIST(root='./data/',
                                   train=False,
                                   transform=transforms.ToTensor())

        # 返回一个数据迭代器
        # shuffle：是否打乱顺序
        train_loader = torch.utils.data.DataLoader(dataset=train_data,
                                                   batch_size=Config.batch_size,
                                                   shuffle=True)

        test_loader = torch.utils.data.DataLoader(dataset=test_data,
                                                  batch_size=Config.batch_size,
                                                  shuffle=False)
        return train_loader, test_loader

    def train_step(self):
        steps = 0
        start_time = datetime.now()

        print("Training & Evaluating based on '%s'......" % self.model)
        for epoch in range(Config.epoch):
            print("Epoch {:3}.".format(epoch + 1))

            for data, label in self.train:
                data, label = Variable(data.cpu()), Variable(label.cpu())
                # LSTM输入为3维，CNN输入为4维
                if self.model == "LSTM":
                    data = data.view(-1, 28, 28)
                self.optimizer.zero_grad()  # 将梯度归零
                outputs = self.net(data)  # 将数据传入网络进行前向运算
                loss = self.criterion(outputs, label)  # 得到损失函数
                loss.backward()  # 反向传播
                self.optimizer.step()  # 通过梯度做一步参数更新

                # 每100次打印一次结果
                if steps % Config.print_per_step == 0:
                    _, predicted = torch.max(outputs, 1)
                    correct = int(sum(predicted == label))  # 计算预测正确个数
                    accuracy = correct / Config.batch_size  # 计算准确率
                    end_time = datetime.now()
                    time_diff = (end_time - start_time).seconds
                    time_usage = '{:3}m{:3}s'.format(int(time_diff / 60), time_diff % 60)
                    msg = "Step {:5}, Loss:{:6.2f}, Accuracy:{:8.2%}, Time usage:{:9}."
                    print(msg.format(steps, loss, accuracy, time_usage))

                steps += 1

        test_loss = 0.
        test_correct = 0
        for data, label in self.test:
            data, label = Variable(data.cpu()), Variable(label.cpu())
            if self.model == "LSTM":
                data = data.view(-1, 28, 28)
            outputs = self.net(data)
            loss = self.criterion(outputs, label)
            test_loss += loss * Config.batch_size
            _, predicted = torch.max(outputs, 1)
            correct = int(sum(predicted == label))
            test_correct += correct

        accuracy = test_correct / len(self.test.dataset)
        loss = test_loss / len(self.test.dataset)
        print("Test Loss: {:5.2f}, Accuracy: {:6.2%}".format(loss, accuracy))

        end_time = datetime.now()
        time_diff = (end_time - start_time).seconds
        print("Time Usage: {:5.2f} mins.".format(time_diff / 60.))


if __name__ == "__main__":
    p = TrainProcess(model='CNN')
    p.train_step()

Peace~~