2023-11-30-526互联

import torchvision
import torchvision.datasets as datasets
from torch import nn
from torch.utils.data import DataLoader
import torch
# 设置学习率和训练轮数
learning_rate = 1e-3
epoch = 50

# 准备数据集
train_data = datasets.ImageFolder('./data/train', transform=torchvision.transforms.Compose([
                                torchvision.transforms.Resize((32, 32)),
                                torchvision.transforms.ToTensor()
                            ]))
test_data = datasets.ImageFolder('./data/test', transform=torchvision.transforms.Compose([
                                torchvision.transforms.Resize((32, 32)),
                                torchvision.transforms.ToTensor()
                            ]))

train_data_size = len(train_data)
test_data_size = len(test_data)

train_dataloader = DataLoader(train_data, batch_size=64, shuffle=True)
test_dataloader = DataLoader(test_data, batch_size=64, shuffle=True)

# 搭建神经网络
class Tudui(nn.Module):
    def __init__(self):
        super(Tudui, self).__init__()
        self.model = nn.Sequential(
            nn.Conv2d(3, 32, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Conv2d(32, 32, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Conv2d(32, 64, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Conv2d(64, 64, 5, 1, 2),
            nn.MaxPool2d(2),
            nn.Flatten(),
            nn.Linear(64 * 2 * 2, 128),
            nn.Linear(128, 64),
            nn.Linear(64, 5)
        )
        self.dropout = nn.Dropout(0.5)

    def forward(self, x):
        x = self.model(x)
        x = self.dropout(x)
        return x

tudui = Tudui()

loss_fn = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(tudui.parameters(), lr=learning_rate)

total_train_step = 0
total_test_step = 0

for i in range(epoch):
    print("-------第 {} 轮训练开始-------".format(i+1))

    # 训练步骤开始
    tudui.train()
    for data in train_dataloader:
        imgs, targets = data
        outputs = tudui(imgs)
        loss = loss_fn(outputs, targets)

        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        total_train_step += 1
        if total_train_step % 100 == 0:
            print("训练次数：{}, Loss: {}".format(total_train_step, loss.item()))

    # 测试步骤开始
    tudui.eval()
    total_test_loss = 0
    correct_predictions = 0
    total_samples = 0
    with torch.no_grad():
        for data in test_dataloader:
            imgs, targets = data
            outputs = tudui(imgs)
            loss = loss_fn(outputs, targets)
            total_test_loss += loss.item()
            _, predicted = torch.max(outputs, 1)
            correct_predictions += (predicted == targets).sum().item()
            total_samples += targets.size(0)

    print("整体测试集上的Loss: {}".format(total_test_loss))
    print("整体测试集上的正确率: {}".format(correct_predictions / total_samples))
    total_test_step += 1
    torch.save(tudui, "tudui_{}.pth".format(i))
    print("模型已保存")