本文实现了一项基于Torch的CIFAR10图像分类任务，代码主要由模型、训练、展示三部分构成。
Train.py

import torch
import torchvision
from tensorboardX import SummaryWriter
from torch import nn
from torch.utils.data import DataLoader
from model import *

train_data = torchvision.datasets.CIFAR10(root="./data",train=True,
                                          transform=torchvision.transforms.ToTensor(),download=True)
test_data = torchvision.datasets.CIFAR10(root="./data",train=False,
                                         transform=torchvision.transforms.ToTensor(),download=True)

device = torch.device("cuda")

# 查看数据集的长度
train_data_size = len(train_data)
print("训练集的长度为：{}".format(train_data_size))
test_data_size = len(test_data)
print("测试集的长度为：{}".format(test_data_size))

#加载数据
train_dataloader = DataLoader(dataset=train_data,batch_size=64)
test_dataloader = DataLoader(dataset=test_data,batch_size=64)

#创建网络模型
First_try = first_try()
First_try = First_try.to(device)

#定义损失函数
Loss_fn = nn.CrossEntropyLoss()
Loss_fn = Loss_fn.to(device)
#优化器
learning_rate = 1e-2  #0.01
optimizer = torch.optim.SGD(First_try.parameters(),lr=learning_rate)

#设置训练参数：
total_train_step = 0 #记录训练次数
total_test_step = 0 #记录测试次数

epoch = 20  #训练轮数



#添加tensorboard
writer = SummaryWriter("./logs_train")

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

    #训练步骤开始
    First_try.train()
    for data in train_dataloader:

        imgs , targets = data

        imgs = imgs.to(device)
        targets = targets.to(device)

        outputs = First_try(imgs)
        loss = Loss_fn(outputs , targets)

        #优化器优化模型
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()

        total_train_step = total_train_step + 1
        if total_train_step%1000 == 0:
            print("训练次数 : {} , Loss : {}".format(total_train_step,loss.item()))
            writer.add_scalar("train_loss" , loss.item() , total_train_step)


    #测试步骤开始
    First_try.eval()
    total_test_loss = 0
    total_accuracy = 0

    with torch.no_grad():
        for data in test_dataloader:
            imgs , targets = data

            imgs = imgs.to(device)
            targets = targets.to(device)

            outputs = First_try(imgs)
            loss = Loss_fn(outputs,targets)
            total_test_loss = total_test_loss + loss.item()

            #准确率
            accuracy = (outputs.argmax(1) == targets).sum()
            total_accuracy = total_accuracy + accuracy

    print("整体测试集上的loss: {} ".format(total_test_loss))
    print("整体测试集上的ACC: {} ".format(total_accuracy / test_data_size))

    writer.add_scalar("test_loss" , total_test_loss , total_test_step)
    writer.add_scalar("test_ACC", total_accuracy / test_data_size , total_test_step)

    total_test_step = total_test_step + 1

    #保存模型
    torch.save(First_try , "./savemodels/First_try_{}.pth".format(i))
    print("模型已经保持")

writer.close()

model.py

import torch
from torch import nn

#搭建模型
class first_try(nn.Module):
    def __init__(self):
        super().__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__':
    First_try = first_try()
    input = torch.ones((64,3,32,32))
    output = First_try(input)
    print(output.shape)

demo.py

import torchvision
from PIL import Image
from model import *

device = torch.device("cuda")

image_path = "./test_picture.jpg"
image = Image.open(image_path)
print(image)

transform = torchvision.transforms.Compose([
    torchvision.transforms.Resize((32,32)),
    torchvision.transforms.ToTensor()
])

image = transform(image)
print(image.shape)
image = image.to(device)

test_model = first_try()
test_model = test_model.to(device)
test_model = torch.load('./savemodels/First_try_25.pth')


image = torch.reshape(image , (1,3,32,32))

test_model.eval()
with torch.no_grad():
    output = test_model(image)

print(output)
print(output.argmax(1))

查看Tensorboard日志：

tensorboard --logdir=logs_train --port=6030