- 🍨 本文为🔗365天深度学习训练营 中的学习记录博客
- 🍖 原作者：K同学啊
🚀我的环境：
- 语言环境：python 3.12.6
- 编译器：jupyter lab
- 深度学习环境：Pytorch
前期准备
import torch
import torch.nn as nn
import torchvision.transforms as transforms
import torchvision
from torchvision import transforms, datasets
import os,PIL,pathlib,warnings
warnings.filterwarnings("ignore") #忽略警告信息
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
device
device(type='cpu')
import os,PIL,random,pathlib
data_dir = 'd:/Users/yxy/Desktop/PotatoPlants'
data_dir = pathlib.Path(data_dir)
data_paths = list(data_dir.glob('*'))
classeNames = [str(path).split("\\")[5] for path in data_paths]
classeNames
['Early_blight', 'healthy', 'Late_blight', 'PotatoPlants']
# 关于transforms.Compose的更多介绍可以参考：https://blog.youkuaiyun.com/qq_38251616/article/details/124878863
train_transforms = transforms.Compose([
transforms.Resize([224, 224]), # 将输入图片resize成统一尺寸
# transforms.RandomHorizontalFlip(), # 随机水平翻转
transforms.ToTensor(), # 将PIL Image或numpy.ndarray转换为tensor，并归一化到[0,1]之间
transforms.Normalize( # 标准化处理-->转换为标准正太分布（高斯分布），使模型更容易收敛
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]) # 其中 mean=[0.485,0.456,0.406]与std=[0.229,0.224,0.225] 从数据集中随机抽样计算得到的。
])
test_transform = transforms.Compose([
transforms.Resize([224, 224]), # 将输入图片resize成统一尺寸
transforms.ToTensor(), # 将PIL Image或numpy.ndarray转换为tensor，并归一化到[0,1]之间
transforms.Normalize( # 标准化处理-->转换为标准正太分布（高斯分布），使模型更容易收敛
mean=[0.485, 0.456, 0.406],
std=[0.229, 0.224, 0.225]) # 其中 mean=[0.485,0.456,0.406]与std=[0.229,0.224,0.225] 从数据集中随机抽样计算得到的。
])
total_data = datasets.ImageFolder("d:/Users/yxy/Desktop/PotatoPlants",transform=train_transforms)
total_data
Dataset ImageFolder
Number of datapoints: 2152
Root location: d:/Users/yxy/Desktop/PotatoPlants
StandardTransform
Transform: Compose(
Resize(size=[224, 224], interpolation=bilinear, max_size=None, antialias=True)
ToTensor()
Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])
)
total_data.class_to_idx
{'Early_blight': 0, 'Late_blight': 1, 'PotatoPlants': 2, 'healthy': 3}
train_size = int(0.8 * len(total_data))
test_size = len(total_data) - train_size
train_dataset, test_dataset = torch.utils.data.random_split(total_data, [train_size, test_size])
train_dataset, test_dataset
(<torch.utils.data.dataset.Subset at 0x2931200bb30>,
<torch.utils.data.dataset.Subset at 0x29311f39b80>)
batch_size = 32
train_dl = torch.utils.data.DataLoader(train_dataset,
batch_size=batch_size,
shuffle=True,
num_workers=1)
test_dl = torch.utils.data.DataLoader(test_dataset,
batch_size=batch_size,
shuffle=True,
num_workers=1)
for X, y in test_dl:
print("Shape of X [N, C, H, W]: ", X.shape)
print("Shape of y: ", y.shape, y.dtype)
break
Shape of X [N, C, H, W]: torch.Size([32, 3, 224, 224])
Shape of y: torch.Size([32]) torch.int64
构建VGG
import torch
import torch.nn as nn
class vgg16(nn.Module):
def __init__(self, num_classes=4):
super(vgg16, self).__init__()
self.block1 = nn.Sequential(
nn.Conv2d(3, 64, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(64, 64, kernel_size=3, padding=1), nn.ReLU(),
nn.MaxPool2d(kernel_size=2, stride=2)
)
self.block2 = nn.Sequential(
nn.Conv2d(64, 128, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(128, 128, kernel_size=3, padding=1), nn.ReLU(),
nn.MaxPool2d(kernel_size=2, stride=2)
)
self.block3 = nn.Sequential(
nn.Conv2d(128, 256, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(256, 256, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(256, 256, kernel_size=3, padding=1), nn.ReLU(),
nn.MaxPool2d(kernel_size=2, stride=2)
)
self.block4 = nn.Sequential(
nn.Conv2d(256, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(512, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(512, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.MaxPool2d(kernel_size=2, stride=2)
)
self.block5 = nn.Sequential(
nn.Conv2d(512, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(512, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.Conv2d(512, 512, kernel_size=3, padding=1), nn.ReLU(),
nn.MaxPool2d(kernel_size=2, stride=2)
)
# 使用 AdaptiveAvgPool2d 确保特征图是 (7,7)
self.avgpool = nn.AdaptiveAvgPool2d((7, 7))
self.classifier = nn.Sequential(
nn.Linear(512 * 7 * 7, 4096), nn.ReLU(),
nn.Linear(4096, 4096), nn.ReLU(),
nn.Linear(4096, num_classes) # num_classes = 4
)
def forward(self, x):
x = self.block1(x)
x = self.block2(x)
x = self.block3(x)
x = self.block4(x)
x = self.block5(x)
x = self.avgpool(x) # 确保 (7,7)
x = torch.flatten(x, start_dim=1)
x = self.classifier(x)
return x
device = "cuda" if torch.cuda.is_available() else "cpu"
print("Using {} device".format(device))
model = vgg16().to(device)
print(model)
Using cpu device
vgg16(
(block1): Sequential(
(0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU()
(4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(block2): Sequential(
(0): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU()
(4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(block3): Sequential(
(0): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU()
(4): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(5): ReLU()
(6): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(block4): Sequential(
(0): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU()
(4): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(5): ReLU()
(6): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(block5): Sequential(
(0): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(1): ReLU()
(2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(3): ReLU()
(4): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))
(5): ReLU()
(6): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)
)
(avgpool): AdaptiveAvgPool2d(output_size=(7, 7))
(classifier): Sequential(
(0): Linear(in_features=25088, out_features=4096, bias=True)
(1): ReLU()
(2): Linear(in_features=4096, out_features=4096, bias=True)
(3): ReLU()
(4): Linear(in_features=4096, out_features=4, bias=True)
)
)
# 统计模型参数量以及其他指标
import torchsummary as summary
summary.summary(model, (3, 224, 224))
----------------------------------------------------------------
Layer (type) Output Shape Param #
================================================================
Conv2d-1 [-1, 64, 224, 224] 1,792
ReLU-2 [-1, 64, 224, 224] 0
Conv2d-3 [-1, 64, 224, 224] 36,928
ReLU-4 [-1, 64, 224, 224] 0
MaxPool2d-5 [-1, 64, 112, 112] 0
Conv2d-6 [-1, 128, 112, 112] 73,856
ReLU-7 [-1, 128, 112, 112] 0
Conv2d-8 [-1, 128, 112, 112] 147,584
ReLU-9 [-1, 128, 112, 112] 0
MaxPool2d-10 [-1, 128, 56, 56] 0
Conv2d-11 [-1, 256, 56, 56] 295,168
ReLU-12 [-1, 256, 56, 56] 0
Conv2d-13 [-1, 256, 56, 56] 590,080
ReLU-14 [-1, 256, 56, 56] 0
Conv2d-15 [-1, 256, 56, 56] 590,080
ReLU-16 [-1, 256, 56, 56] 0
MaxPool2d-17 [-1, 256, 28, 28] 0
Conv2d-18 [-1, 512, 28, 28] 1,180,160
ReLU-19 [-1, 512, 28, 28] 0
Conv2d-20 [-1, 512, 28, 28] 2,359,808
ReLU-21 [-1, 512, 28, 28] 0
Conv2d-22 [-1, 512, 28, 28] 2,359,808
ReLU-23 [-1, 512, 28, 28] 0
MaxPool2d-24 [-1, 512, 14, 14] 0
Conv2d-25 [-1, 512, 14, 14] 2,359,808
ReLU-26 [-1, 512, 14, 14] 0
Conv2d-27 [-1, 512, 14, 14] 2,359,808
ReLU-28 [-1, 512, 14, 14] 0
Conv2d-29 [-1, 512, 14, 14] 2,359,808
ReLU-30 [-1, 512, 14, 14] 0
MaxPool2d-31 [-1, 512, 7, 7] 0
AdaptiveAvgPool2d-32 [-1, 512, 7, 7] 0
Linear-33 [-1, 4096] 102,764,544
ReLU-34 [-1, 4096] 0
Linear-35 [-1, 4096] 16,781,312
ReLU-36 [-1, 4096] 0
Linear-37 [-1, 4] 16,388
================================================================
Total params: 134,276,932
Trainable params: 134,276,932
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 0.57
Forward/backward pass size (MB): 218.71
Params size (MB): 512.23
Estimated Total Size (MB): 731.51
----------------------------------------------------------------
训练模型
# 训练循环
def train(dataloader, model, loss_fn, optimizer):
size = len(dataloader.dataset) # 训练集的大小
num_batches = len(dataloader) # 批次数目, (size/batch_size，向上取整)
train_loss, train_acc = 0, 0 # 初始化训练损失和正确率
for X, y in dataloader: # 获取图片及其标签
X, y = X.to(device), y.to(device)
# 计算预测误差
pred = model(X) # 网络输出
loss = loss_fn(pred, y) # 计算网络输出和真实值之间的差距，targets为真实值，计算二者差值即为损失
# 反向传播
optimizer.zero_grad() # grad属性归零
loss.backward() # 反向传播
optimizer.step() # 每一步自动更新
# 记录acc与loss
train_acc += (pred.argmax(1) == y).type(torch.float).sum().item()
train_loss += loss.item()
train_acc /= size
train_loss /= num_batches
return train_acc, train_loss
def test (dataloader, model, loss_fn):
size = len(dataloader.dataset) # 测试集的大小
num_batches = len(dataloader) # 批次数目, (size/batch_size，向上取整)
test_loss, test_acc = 0, 0
# 当不进行训练时，停止梯度更新，节省计算内存消耗
with torch.no_grad():
for imgs, target in dataloader:
imgs, target = imgs.to(device), target.to(device)
# 计算loss
target_pred = model(imgs)
loss = loss_fn(target_pred, target)
test_loss += loss.item()
test_acc += (target_pred.argmax(1) == target).type(torch.float).sum().item()
test_acc /= size
test_loss /= num_batches
return test_acc, test_loss
import copy
optimizer = torch.optim.Adam(model.parameters(), lr= 1e-4)
loss_fn = nn.CrossEntropyLoss() # 创建损失函数
epochs = 40
train_loss = []
train_acc = []
test_loss = []
test_acc = []
best_acc = 0 # 设置一个最佳准确率，作为最佳模型的判别指标
best_model = None # 存储最佳模型
for epoch in range(epochs):
model.train()
epoch_train_acc, epoch_train_loss = train(train_dl, model, loss_fn, optimizer)
model.eval()
epoch_test_acc, epoch_test_loss = test(test_dl, model, loss_fn)
# 保存最佳模型到 best_model
if epoch_test_acc > best_acc:
best_acc = epoch_test_acc
best_model = copy.deepcopy(model)
train_acc.append(epoch_train_acc)
train_loss.append(epoch_train_loss)
test_acc.append(epoch_test_acc)
test_loss.append(epoch_test_loss)
# 获取当前的学习率
lr = optimizer.state_dict()['param_groups'][0]['lr']
template = ('Epoch:{:2d}, Train_acc:{:.1f}%, Train_loss:{:.3f}, Test_acc:{:.1f}%, Test_loss:{:.3f}, Lr:{:.2E}')
print(template.format(epoch+1, epoch_train_acc*100, epoch_train_loss,
epoch_test_acc*100, epoch_test_loss, lr))
# 保存最佳模型到文件中
PATH = 'd:/Users/yxy/Desktop/test_model.pth' # 保存的参数文件名
torch.save(model.state_dict(), PATH)
print('Done')
Epoch: 1, Train_acc:46.5%, Train_loss:1.193, Test_acc:44.1%, Test_loss:1.024, Lr:1.00E-04
Epoch: 2, Train_acc:59.8%, Train_loss:0.858, Test_acc:67.5%, Test_loss:0.766, Lr:1.00E-04
Epoch: 3, Train_acc:66.4%, Train_loss:0.760, Test_acc:70.8%, Test_loss:0.684, Lr:1.00E-04
Epoch: 4, Train_acc:68.2%, Train_loss:0.678, Test_acc:72.4%, Test_loss:0.608, Lr:1.00E-04
Epoch: 5, Train_acc:68.3%, Train_loss:0.640, Test_acc:65.7%, Test_loss:0.621, Lr:1.00E-04
Epoch: 6, Train_acc:69.2%, Train_loss:0.610, Test_acc:63.8%, Test_loss:0.562, Lr:1.00E-04
Epoch: 7, Train_acc:70.7%, Train_loss:0.533, Test_acc:76.3%, Test_loss:0.477, Lr:1.00E-04
Epoch: 8, Train_acc:71.2%, Train_loss:0.569, Test_acc:67.7%, Test_loss:0.511, Lr:1.00E-04
Epoch: 9, Train_acc:73.5%, Train_loss:0.493, Test_acc:69.1%, Test_loss:0.655, Lr:1.00E-04
Epoch:10, Train_acc:73.4%, Train_loss:0.486, Test_acc:75.2%, Test_loss:0.449, Lr:1.00E-04
Epoch:11, Train_acc:74.4%, Train_loss:0.447, Test_acc:79.4%, Test_loss:0.400, Lr:1.00E-04
Epoch:12, Train_acc:75.9%, Train_loss:0.424, Test_acc:69.1%, Test_loss:0.444, Lr:1.00E-04
Epoch:13, Train_acc:75.2%, Train_loss:0.418, Test_acc:68.4%, Test_loss:0.462, Lr:1.00E-04
Epoch:14, Train_acc:75.0%, Train_loss:0.424, Test_acc:71.9%, Test_loss:0.390, Lr:1.00E-04
Epoch:15, Train_acc:75.0%, Train_loss:0.426, Test_acc:75.4%, Test_loss:0.491, Lr:1.00E-04
Epoch:16, Train_acc:76.4%, Train_loss:0.412, Test_acc:77.0%, Test_loss:0.373, Lr:1.00E-04
Epoch:17, Train_acc:77.1%, Train_loss:0.379, Test_acc:80.0%, Test_loss:0.368, Lr:1.00E-04
Epoch:18, Train_acc:78.0%, Train_loss:0.375, Test_acc:80.7%, Test_loss:0.367, Lr:1.00E-04
Epoch:19, Train_acc:78.2%, Train_loss:0.383, Test_acc:70.5%, Test_loss:0.408, Lr:1.00E-04
Epoch:20, Train_acc:78.7%, Train_loss:0.352, Test_acc:78.4%, Test_loss:0.352, Lr:1.00E-04
Epoch:21, Train_acc:79.9%, Train_loss:0.344, Test_acc:80.0%, Test_loss:0.353, Lr:1.00E-04
Epoch:22, Train_acc:79.1%, Train_loss:0.360, Test_acc:79.4%, Test_loss:0.379, Lr:1.00E-04
Epoch:23, Train_acc:80.5%, Train_loss:0.342, Test_acc:78.2%, Test_loss:0.363, Lr:1.00E-04
Epoch:24, Train_acc:82.2%, Train_loss:0.321, Test_acc:77.7%, Test_loss:0.373, Lr:1.00E-04
Epoch:25, Train_acc:82.7%, Train_loss:0.312, Test_acc:78.7%, Test_loss:0.389, Lr:1.00E-04
Epoch:26, Train_acc:81.6%, Train_loss:0.329, Test_acc:73.8%, Test_loss:0.428, Lr:1.00E-04
Epoch:27, Train_acc:82.6%, Train_loss:0.314, Test_acc:72.9%, Test_loss:0.449, Lr:1.00E-04
Epoch:28, Train_acc:85.1%, Train_loss:0.276, Test_acc:76.1%, Test_loss:0.445, Lr:1.00E-04
Epoch:29, Train_acc:87.0%, Train_loss:0.250, Test_acc:78.7%, Test_loss:0.445, Lr:1.00E-04
Epoch:30, Train_acc:88.8%, Train_loss:0.258, Test_acc:75.6%, Test_loss:0.567, Lr:1.00E-04
Epoch:31, Train_acc:89.5%, Train_loss:0.230, Test_acc:76.8%, Test_loss:0.438, Lr:1.00E-04
Epoch:32, Train_acc:91.2%, Train_loss:0.200, Test_acc:74.7%, Test_loss:0.491, Lr:1.00E-04
Epoch:33, Train_acc:88.9%, Train_loss:0.303, Test_acc:70.3%, Test_loss:0.615, Lr:1.00E-04
Epoch:34, Train_acc:87.5%, Train_loss:0.314, Test_acc:75.2%, Test_loss:0.616, Lr:1.00E-04
Epoch:35, Train_acc:95.7%, Train_loss:0.113, Test_acc:78.9%, Test_loss:0.725, Lr:1.00E-04
Epoch:36, Train_acc:93.8%, Train_loss:0.162, Test_acc:76.6%, Test_loss:0.637, Lr:1.00E-04
Epoch:37, Train_acc:94.8%, Train_loss:0.135, Test_acc:73.8%, Test_loss:0.638, Lr:1.00E-04
Epoch:38, Train_acc:98.1%, Train_loss:0.059, Test_acc:74.5%, Test_loss:1.233, Lr:1.00E-04
Epoch:39, Train_acc:85.5%, Train_loss:0.564, Test_acc:77.5%, Test_loss:0.536, Lr:1.00E-04
Epoch:40, Train_acc:85.0%, Train_loss:0.367, Test_acc:73.3%, Test_loss:0.590, Lr:1.00E-04
Done
import matplotlib.pyplot as plt
#隐藏警告
import warnings
warnings.filterwarnings("ignore") #忽略警告信息
plt.rcParams['font.sans-serif'] = ['SimHei'] # 用来正常显示中文标签
plt.rcParams['axes.unicode_minus'] = False # 用来正常显示负号
plt.rcParams['figure.dpi'] = 100 #分辨率
from datetime import datetime
current_time = datetime.now() # 获取当前时间
epochs_range = range(epochs)
plt.figure(figsize=(12, 3))
plt.subplot(1, 2, 1)
plt.plot(epochs_range, train_acc, label='Training Accuracy')
plt.plot(epochs_range, test_acc, label='Test Accuracy')
plt.legend(loc='lower right')
plt.title('Training and Validation Accuracy')
plt.xlabel(current_time) # 打卡请带上时间戳，否则代码截图无效
plt.subplot(1, 2, 2)
plt.plot(epochs_range, train_loss, label='Training Loss')
plt.plot(epochs_range, test_loss, label='Test Loss')
plt.legend(loc='upper right')
plt.title('Training and Validation Loss')
plt.show()
from PIL import Image
classes = list(total_data.class_to_idx)
def predict_one_image(image_path, model, transform, classes):
test_img = Image.open(image_path).convert('RGB')
plt.imshow(test_img) # 展示预测的图片
test_img = transform(test_img)
img = test_img.to(device).unsqueeze(0)
model.eval()
output = model(img)
_,pred = torch.max(output,1)
pred_class = classes[pred]
print(f'预测结果是：{pred_class}')
# 预测训练集中的某张照片
predict_one_image(image_path='d:/Users/yxy/Desktop/PotatoPlants/Early_blight/1.JPG',
model=model,
transform=train_transforms,
classes=classes)
预测结果是：Early_blight
best_model.eval()
epoch_test_acc, epoch_test_loss = test(test_dl, best_model, loss_fn)
epoch_test_acc, epoch_test_loss
(0.8074245939675174, 0.36957609547036036)
总结
使用 torchsummary，可以查看模型的参数量以及相关指标，查看到各层的参数数量、输出形状、总参数量、可训练参数 。
扫一扫
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
