学习资源来自b站，一点点手敲代码初步接触深度学习训练模型。感觉还是很神奇的！！
将训练资源下载下来并通过训练模型来实现，本篇主要用来记录当时的一些代码和注释，方便后续回顾。
####################################### net.py ########################################
import torch
from torch import nn
# 定义一个网络模型
class MyLeNet5(nn.Module):
# 初始化网络
# 主要是复现LeNet-5
def __init__(self):
super(MyLeNet5, self).__init__()
# 卷积层c1
self.c1 = nn.Conv2d(in_channels=1, out_channels=6, kernel_size=5, padding=2)
# 单纯单通道 so in=1，输出为6，约定俗成的卷积核是5，padding可以用公式算出来设置为2
# 激活函数
self.Sigmoid = nn.Sigmoid()
# 平均池化（定义一个池化层） !注意! 池化层不改变通道大小，但是会改变特征图片的窗口大小
self.s2 = nn.AvgPool2d(kernel_size=2, stride=2)
# 卷积核为2，步长为2
# 卷积层c3
self.c3 = nn.Conv2d(in_channels=6, out_channels=16, kernel_size=5)
# 池化层s4
self.s4 = nn.AvgPool2d(kernel_size=2, stride=2)
# 卷积层c5
self.c5 = nn.Conv2d(in_channels=16, out_channels=120, kernel_size=5)
# 平展层
self.flatten = nn.Flatten()
# 设置线性连接层
self.f6 = nn.Linear(120, 84)
# 输入、输出
self.output = nn.Linear(84, 10)
def forward(self, x):
# 用Sigmoid函数激活
x = self.Sigmoid(self.c1(x))
# 池化层
x = self.s2(x)
# 以此类推
x = self.Sigmoid(self.c3(x))
x = self.s4(x)
x = self.c5(x)
x = self.flatten(x)
x = self.f6(x)
x = self.output(x)
return x
if __name__ == "__main__":
# 随机生成一个 批次1，通道1，大小是28*28 实例化
x = torch.rand([1, 1, 28, 28])
model = MyLeNet5()
y = model(x)
######################################## test.py ########################################
import torch
from net import MyLeNet5
from torch.autograd import Variable
from torchvision import datasets, transforms
from torchvision.transforms import ToPILImage
# 将数据转化为tensor格式（数据是矩阵格式，要进行转化为tensor格式）
data_transform = transforms.Compose([
transforms.ToTensor()
])
# 加载训练数据集
train_dataset = datasets.MNIST(root='./data', train=True, transform=data_transform, download=True)
train_dataloader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=16, shuffle=True)
# 加载测试数据集
test_dataset = datasets.MNIST(root='./data', train=False, transform=data_transform, download=True)
test_dataloader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=16, shuffle=True)
# 如果有显卡，转到GPU
device = "cuda" if torch.cuda.is_available() else 'cpu'
# 调用net里面定义的模型，将模型数据转到GPU
model = MyLeNet5().to(device)
model.load_state_dict(torch.load("C:/Users/79926/PycharmProjects/pythonProject1/save_model/best_model.pth"))
# 获取结果
classes = [
"0",
"1",
"2",
"3",
"4",
"5",
"6",
"7",
"8",
"9",
]
# 把tensor转化为图片，方便可视化
show = ToPILImage()
# 进入验证
for i in range(5):
X, y = test_dataset[i][0], test_dataset[i][1]
show(X).show()
# 这里会显示出5张图片
X = Variable(torch.unsqueeze(X, dim=0).float(), requires_grad=False).to(device)
with torch.no_grad():
pred = model(X)
predicted, actual = classes[torch.argmax(pred[0])], classes[y]
print(f'predicted:"{predicted}",actual:"{actual}"')
######################################## train.py ########################################
import torch
from torch import nn
from net import MyLeNet5
from torch.optim import lr_scheduler
from torchvision import datasets, transforms
import os
# 将数据转化为tensor格式（数据是矩阵格式，要进行转化为tensor格式）
data_transform = transforms.Compose([
transforms.ToTensor()
])
# 加载训练数据集
train_dataset = datasets.MNIST(root='./data', train=True, transform=data_transform, download=True)
train_dataloader = torch.utils.data.DataLoader(dataset=train_dataset, batch_size=16, shuffle=True)
# 加载测试数据集
test_dataset = datasets.MNIST(root='./data', train=False, transform=data_transform, download=True)
test_dataloader = torch.utils.data.DataLoader(dataset=test_dataset, batch_size=16, shuffle=True)
# 如果有显卡，转到GPU
device = "cuda" if torch.cuda.is_available() else 'cpu'
# 调用net里面定义的模型，将模型数据转到GPU
model = MyLeNet5().to(device)
# 定义一个损失函数（交叉熵损失）
loss_fn = nn.CrossEntropyLoss()
# 定义一个优化器
# (梯度下降)
optimizer = torch.optim.SGD(model.parameters(), lr=1e-3, momentum=0.9)
# 学习率每隔10轮变为原来的0.1
lr_scheduler = lr_scheduler.StepLR(optimizer, step_size=10, gamma=0.1)
# 定义训练函数
def train(dataloader, model, loss_fn, optimizer):
loss, current, n = 0.0, 0.0, 0
for batch, (X, y) in enumerate(dataloader):
# 前向传播
X, y = X.to(device), y.to(device)
output = model(X)
# 损失函数（用来反向传播）
cur_loss = loss_fn(output, y)
_, pred = torch.max(output, axis=1)
# 计算精确度（累加->一轮的）
cur_acc = torch.sum(y == pred) / output.shape[0]
optimizer.zero_grad()
cur_loss.backward()
optimizer.step()
loss += cur_loss.item()
current += cur_acc.item()
n = n + 1
print("train_loss" + str(loss / n))
print("train_acc" + str(current / n))
def val(dataloader, model, loss_fn):
model.eval()
loss, current, n = 0.0, 0.0, 0
with torch.no_grad():
for batch, (X, y) in enumerate(dataloader):
# 前向传播
X, y = X.to(device), y.to(device)
output = model(X)
# 损失函数（用来反向传播）
cur_loss = loss_fn(output, y)
_, pred = torch.max(output, axis=1)
cur_acc = torch.sum(y == pred) / output.shape[0]
loss += cur_loss.item()
current += cur_acc.item()
n = n + 1
print("val_loss" + str(loss / n))
print("val_acc" + str(current / n))
return current/n
# 开始训练
epoch = 50
min_acc = 0
for t in range(epoch):
print(f'epoch{t + 1}\n--------------')
train(train_dataloader, model, loss_fn, optimizer)
a=val(test_dataloader, model, loss_fn)
#保存最好模型权重
if a>min_acc:
folder = 'save_model'
if not os.path.exists(folder):
os.mkdir('save_model')
min_acc = a
print('save best model')
torch.save(model.state_dict(),'save_model/best_model.pth')
print('Done!')
附：该up主视频资源：（讲的很棒）
网络模型搭建_哔哩哔哩_bilibili
训练模型搭建_哔哩哔哩_bilibili
测试模型搭建_哔哩哔哩_bilibili
