记一次模型训练CIFAR10

Introduction

采用数据集CIFAR10，网络模型如下：

Structure-of-CIFAR10-quick-model

Code

代码如下，分别是网络模型和训练代码

from torch import nn
import torch
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.Flatten(),
            nn.Linear(64 * 4 * 4, 64),
            nn.Linear(64, 10)
        )
    def forward(self, x):
        x = self.model(x)
        return x
if __name__ == '__main__':
    tudui = Tudui()
    input = torch.ones((64, 3, 32, 32))
    output = tudui(input)
    print(output.shape)

训练代码：

import torchvision
import torch
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
from model import *
from torch import nn
from torch.nn import Module
train_data = torchvision.datasets.CIFAR10("CIFAR10", train=True, transform=torchvision.transforms.ToTensor(), download=True)
test_data = torchvision.datasets.CIFAR10("CIFAR10", train=False, transform=torchvision.transforms.ToTensor(), download=True)
train_data_size = len(train_data)
test_data_size = len(test_data)
train_dataloader = DataLoader(train_data, batch_size=64)
test_dataloader = DataLoader(test_data, batch_size=64)
# create model
tudui = Tudui()
# create loss function
loss_fn = nn.CrossEntropyLoss()
# optim
learning_rate = 0.01
optimizer = torch.optim.SGD(tudui.parameters(), lr=learning_rate)
# set some parameters to train
total_train_step = 0
total_test_step = 0
epoch = 10
# add tensorboard
writer = SummaryWriter("logs_train")
# for i in range(epoch):
i = 0
while True:
    print("-------- This is No. {} times of training ----------".format(i + 1))
    for data in train_dataloader:
        imgs, targets = data
        outputs = tudui(imgs)
        loss = loss_fn(outputs, targets)
        # do some optimization
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
        total_train_step += 1
        if total_train_step % 100 == 0:
            print("Times of training {}, loss: {}".format(total_train_step, loss.item()))
            writer.add_scalar("train_loss", loss.item(), total_train_step)
    # test the result of training
    total_test_loss = 0
    total_accuracy = 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()
            accuracy = (outputs.argmax(1) == targets).sum()
            total_accuracy += accuracy
    print("Loss on the whole test set: {} ".format(total_test_loss))
    print("Accuracy rate on the whole test set: {} ".format(total_accuracy / test_data_size))
    if total_accuracy / test_data_size > 0.8:
        break
    writer.add_scalar("test_loss", total_test_loss, total_test_step)
    writer.add_scalar("test_accuracy", total_accuracy / test_data_size, total_test_step)
    total_test_step += 1

    torch.save(tudui, "tudui_{}.pth".format(i))
    print("Model has been saved ")
    i = i + 1
writer.close()