继续之前的学习
过拟合的判断
import torch
import torch.nn as nn
import torch.optim as optim
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
import time
import matplotlib.pyplot as plt
from tqdm import tqdm
import warnings
warnings.filterwarnings("ignore")
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print(f"使用设备: {device}")
iris = load_iris()
X = iris.data
y = iris.target
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
scaler = MinMaxScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)
X_train = torch.FloatTensor(X_train).to(device)
y_train = torch.LongTensor(y_train).to(device)
X_test = torch.FloatTensor(X_test).to(device)
y_test = torch.LongTensor(y_test).to(device)
class MLP(nn.Module):
def __init__(self):
super(MLP, self).__init__()
self.fc1 = nn.Linear(4, 10)
self.relu = nn.ReLU()
self.fc2 = nn.Linear(10, 3)
def forward(self, x):
out = self.fc1(x)
out = self.relu(out)
out = self.fc2(out)
return out
model = MLP().to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.SGD(model.parameters(), lr=0.01)
num_epochs = 20000
# 用于存储每200个epoch的损失值和对应的epoch数
train_losses = []
test_losses = []
epochs = []
start_time = time.time()
with tqdm(total=num_epochs, desc="训练进度", unit="epoch") as pbar:
for epoch in range(num_epochs):
outputs = model(X_train)
train_loss = criterion(outputs, y_train)
optimizer.zero_grad()
train_loss.backward()
optimizer.step()
if (epoch + 1) % 200 == 0:
model.eval()
with torch.no_grad():
test_outputs = model(X_test)
test_loss = criterion(test_outputs, y_test)
model.train()
train_losses.append(train_loss.item())
test_losses.append(test_loss.item())
epochs.append(epoch + 1)
pbar.set_postfix({'Train Loss': f'{train_loss.item():.4f}', 'Test Loss': f'{test_loss.item():.4f}'})
if (epoch + 1) % 1000 == 0:
pbar.update(1000)
if pbar.n < num_epochs:
pbar.update(num_epochs - pbar.n)
time_all = time.time() - start_time
print(f'Training time: {time_all:.2f} seconds')
plt.figure(figsize=(10, 6))
plt.plot(epochs, train_losses, label='Train Loss')
plt.plot(epochs, test_losses, label='Test Loss')
plt.xlabel('Epoch')
plt.ylabel('Loss')
plt.title('Training and Test Loss over Epochs')
plt.legend()
plt.grid(True)
plt.show()
model.eval()
with torch.no_grad():
outputs = model(X_test)
_, predicted = torch.max(outputs, 1) # torch.max(outputs, 1)返回每行的最大值和对应的索引
correct = (predicted == y_test).sum().item()
accuracy = correct / y_test.size(0)
print(f'测试集准确率: {accuracy * 100:.2f}%') 模型的保存和加载
深度学习中模型的保存与加载主要涉及参数（权重）和整个模型结构的存储，同时需兼顾训练状态（如优化器参数、轮次等）以支持断点续训。
仅保存模型参数
-原理：保存模型的权重参数，不保存模型结构代码。加载时需提前定义与训练时一致的模型类。
-优点：文件体积小（仅含参数），跨框架兼容性强（需自行定义模型结构）
早停法(early stop)
我们梳理下过拟合的情况
正常情况：训练集和测试集损失同步下降，最终趋于稳定
过拟合：训练集损失持续下降，但测试集损失在某一时刻开始上升（或不再下降）
如果可以监控验证集的指标不再变好，此时提前终止训练，避免模型对训练集过度拟合。----监控的对象是验证集的指标。这种策略叫早停法。
import torch
import torch.nn as nn
import torch.optim as optim
from sklearn.datasets import load_iris
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MinMaxScaler
import time
import matplotlib.pyplot as plt
from tqdm import tqdm # 导入tqdm库用于进度条显示
import warnings
warnings.filterwarnings("ignore") # 忽略警告信息
# 设置GPU设备
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
print(f"使用设备: {device}")
# 加载鸢尾花数据集
iris = load_iris()
X = iris.data # 特征数据
y = iris.target # 标签数据
# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
# 归一化数据
scaler = MinMaxScaler()
X_train = scaler.fit_transform(X_train)
X_test = scaler.transform(X_test)
# 将数据转换为PyTorch张量并移至GPU
X_train = torch.FloatTensor(X_train).to(device)
y_train = torch.LongTensor(y_train).to(device)
X_test = torch.FloatTensor(X_test).to(device)
y_test = torch.LongTensor(y_test).to(device)
class MLP(nn.Module):
def __init__(self):
super(MLP, self).__init__()
self.fc1 = nn.Linear(4, 10) # 输入层到隐藏层
self.relu = nn.ReLU()
self.fc2 = nn.Linear(10, 3) # 隐藏层到输出层
def forward(self, x):
out = self.fc1(x)
out = self.relu(out)
out = self.fc2(out)
return out
# 实例化模型并移至GPU
model = MLP().to(device)
# 分类问题使用交叉熵损失函数
criterion = nn.CrossEntropyLoss()
# 使用随机梯度下降优化器
optimizer = optim.SGD(model.parameters(), lr=0.01)
# 训练模型
num_epochs = 20000 # 训练的轮数
# 用于存储每200个epoch的损失值和对应的epoch数
train_losses = [] # 存储训练集损失
test_losses = [] # 存储测试集损失
epochs = []
# ===== 新增早停相关参数 =====
best_test_loss = float('inf') # 记录最佳测试集损失
best_epoch = 0 # 记录最佳epoch
patience = 50 # 早停耐心值（连续多少轮测试集损失未改善时停止训练）
counter = 0 # 早停计数器
early_stopped = False # 是否早停标志
# ==========================
start_time = time.time() # 记录开始时间
# 创建tqdm进度条
with tqdm(total=num_epochs, desc="训练进度", unit="epoch") as pbar:
# 训练模型
for epoch in range(num_epochs):
# 前向传播
outputs = model(X_train) # 隐式调用forward函数
train_loss = criterion(outputs, y_train)
# 反向传播和优化
optimizer.zero_grad()
train_loss.backward()
optimizer.step()
# 记录损失值并更新进度条
if (epoch + 1) % 200 == 0:
# 计算测试集损失
model.eval()
with torch.no_grad():
test_outputs = model(X_test)
test_loss = criterion(test_outputs, y_test)
model.train()
train_losses.append(train_loss.item())
test_losses.append(test_loss.item())
epochs.append(epoch + 1)
# 更新进度条的描述信息
pbar.set_postfix({'Train Loss': f'{train_loss.item():.4f}', 'Test Loss': f'{test_loss.item():.4f}'})
# ===== 新增早停逻辑 =====
if test_loss.item() < best_test_loss: # 如果当前测试集损失小于最佳损失
best_test_loss = test_loss.item() # 更新最佳损失
best_epoch = epoch + 1 # 更新最佳epoch
counter = 0 # 重置计数器
# 保存最佳模型
torch.save(model.state_dict(), 'best_model.pth')
else:
counter += 1
if counter >= patience:
print(f"早停触发！在第{epoch+1}轮，测试集损失已有{patience}轮未改善。")
print(f"最佳测试集损失出现在第{best_epoch}轮，损失值为{best_test_loss:.4f}")
early_stopped = True
break # 终止训练循环
# ======================
# 每1000个epoch更新一次进度条
if (epoch + 1) % 1000 == 0:
pbar.update(1000) # 更新进度条
# 确保进度条达到100%
if pbar.n < num_epochs:
pbar.update(num_epochs - pbar.n) # 计算剩余的进度并更新
time_all = time.time() - start_time # 计算训练时间
print(f'Training time: {time_all:.2f} seconds')
# ===== 新增：加载最佳模型用于最终评估 =====
if early_stopped:
print(f"加载第{best_epoch}轮的最佳模型进行最终评估...")
model.load_state_dict(torch.load('best_model.pth'))
# ================================
# 可视化损失曲线
plt.figure(figsize=(10, 6))
plt.plot(epochs, train_losses, label='Train Loss')
plt.plot(epochs, test_losses, label='Test Loss')
plt.xlabel('Epoch')
plt.ylabel('Loss')
plt.title('Training and Test Loss over Epochs')
plt.legend()
plt.grid(True)
plt.show()
# 在测试集上评估模型
model.eval()
with torch.no_grad():
outputs = model(X_test)
_, predicted = torch.max(outputs, 1)
correct = (predicted == y_test).sum().item()
accuracy = correct / y_test.size(0)
print(f'测试集准确率: {accuracy * 100:.2f}%') 
