pytorch深度学习小土堆的训练模型

1.将建立的深度学习的模型放入一个py文件中,在mian函数中进行全1的模型检测

import torch
from torch import nn
from torch.nn import Conv2d, MaxPool2d, Flatten,Linear,Sequential

class Tudui(nn.Module):
    def __init__(self):
        super(Tudui, self).__init__()
        self.model1 = Sequential(
            Conv2d(3, 32, 5, padding=2),
            MaxPool2d(2),
            Conv2d(32, 32, 5, padding=2),
            MaxPool2d(2),
            Conv2d(32, 64, 5, padding=2),
            MaxPool2d(2),
            Flatten(),
            Linear(1024, 64),
            Linear(64, 10)
        )
    def forward(self,x):
        x = self.model1(x)
        return x

if __name__ == '__main__':
    tudui = Tudui()
    input = torch.ones((64,3,32,32))
    output = tudui(input)
    print(output.shape)

2.模板大概流程:(1)读取训练集,测试集数据,获取数据总数量

                            (2)建立训练集,测试集loader

                            (3)引入网络模型

                            (4)设置损失函数

                            (5)建立优化器

                            (6)设置训练轮数后开始训练和测试。训练时先将数据放入网络得到output,调用loss函数,将优化器梯度归零,loss函数反向传播,返回梯度,优化器根据梯度进行训练,最终显示loss;测试时首先要在梯度为0的情况下开始进行测试,同样得到output,调用loss函数,可以测试正确率,最终显示总测试集的loss。期间可以用tensorboard将数据可视化。

                            (7)保存模型,每轮换进行一次保存

import torch
from torch import nn
import torchvision
from model import Tudui
from torch.utils.data import DataLoader
from torch.utils.tensorboard import SummaryWriter
writer = SummaryWriter("train")
test_data = torchvision.datasets.CIFAR10("./dataset0",False,torchvision.transforms.ToTensor(),
                                       download=True)
train_data = torchvision.datasets.CIFAR10("./dataset0",True,torchvision.transforms.ToTensor(),
                                          download=True)
test_loader = DataLoader(test_data,64)
train_loader = DataLoader(train_data,64)

test_data_size = len(test_data)

#创建网络模型
tudui = Tudui()

#损失函数
loss_fn = nn.CrossEntropyLoss()

#优化器
learn_rate = 1e-2
optim = torch.optim.SGD(tudui.parameters(),lr=learn_rate)
#训练的参数
#训练的次数
total_train_step = 0
#测试的次数
total_test_step = 0
#训练的轮数
epoch = 10

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

    #开始训练
    tudui.train()
    for data in train_loader:
        imgs,target = data
        outputs = tudui(imgs)
        loss = loss_fn(outputs,target)
        #优化器优化模型
        optim.zero_grad()
        loss.backward()
        optim.step()

        total_train_step = total_train_step+1
        if total_train_step%100 == 0:
            print("训练次数为:{},loss:{}".format(total_train_step,loss.item()))
            writer.add_scalar("train_loss",loss.item(),total_train_step)
    #测试步骤
    tudui.eval()
    total_test_loss = 0
    total_accuracy = 0
    with torch.no_grad():
        for data in test_loader:
            imgs,targets = data
            output = tudui(imgs)
            loss = loss_fn(output,targets)
            total_test_loss = total_test_loss+loss.item()
            accurracy = (output.argmax(1)==targets).sum()
            total_accuracy = total_accuracy+accurracy
    print("整体测试上的loss:{}".format(total_test_loss))
    print("整体测试上的正确率:{}".format(total_accuracy/test_data_size))
    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 = total_test_step+1

    # torch.save(tudui,"tudui{}.pth".format(i))
    # print("模型已经保存")

writer.close()

你可能感兴趣的:(pytorch,深度学习,神经网络)