1. 彩色和灰度圖片測試和訓練的規范寫法：封裝在函數中
2. 展平操作：除第一個維度batchsize外全部展平
3. dropout操作：訓練階段隨機丟棄神經元，測試階段eval模式關閉dropout

作業：仔細學習下測試和訓練代碼的邏輯，這是基礎，這個代碼框架后續會一直沿用，后續的重點慢慢就是轉向模型定義階段了。

import torch
import torch.nn as nn
import torch.optim as optim
from torchvision import datasets, transforms
from torch.utils.data import DataLoader
import matplotlib.pyplot as plt
import numpy as np# 設置中文字體支持
plt.rcParams["font.family"] = ["SimHei"]
plt.rcParams['axes.unicode_minus'] = False  # 解決負號顯示問題# 1. 數據預處理
transform = transforms.Compose([transforms.ToTensor(),  # 轉換為張量并歸一化到[0,1]transforms.Normalize((0.1307,), (0.3081,))  # MNIST數據集的均值和標準差
])# 2. 加載MNIST數據集
train_dataset = datasets.MNIST(root='./data',train=True,download=True,transform=transform
)test_dataset = datasets.MNIST(root='./data',train=False,transform=transform
)# 3. 創建數據加載器
batch_size = 64  # 每批處理64個樣本
train_loader = DataLoader(train_dataset, batch_size=batch_size, shuffle=True)
test_loader = DataLoader(test_dataset, batch_size=batch_size, shuffle=False)# 4. 定義模型、損失函數和優化器
class MLP(nn.Module):def __init__(self):super(MLP, self).__init__()self.flatten = nn.Flatten()  # 將28x28的圖像展平為784維向量self.layer1 = nn.Linear(784, 128)  # 第一層：784個輸入，128個神經元self.relu = nn.ReLU()  # 激活函數self.layer2 = nn.Linear(128, 10)  # 第二層：128個輸入，10個輸出（對應10個數字類別）def forward(self, x):x = self.flatten(x)  # 展平圖像x = self.layer1(x)   # 第一層線性變換x = self.relu(x)     # 應用ReLU激活函數x = self.layer2(x)   # 第二層線性變換，輸出logitsreturn x# 檢查GPU是否可用
device = torch.device("cuda" if torch.cuda.is_available() else "cpu")# 初始化模型
model = MLP()
model = model.to(device)  # 將模型移至GPU（如果可用）criterion = nn.CrossEntropyLoss()  # 交叉熵損失函數，適用于多分類問題
optimizer = optim.Adam(model.parameters(), lr=0.001)  # Adam優化器# 5. 訓練模型（記錄每個 iteration 的損失）
def train(model, train_loader, test_loader, criterion, optimizer, device, epochs):model.train()  # 設置為訓練模式# 新增：記錄每個 iteration 的損失all_iter_losses = []  # 存儲所有 batch 的損失iter_indices = []     # 存儲 iteration 序號（從1開始）for epoch in range(epochs):running_loss = 0.0correct = 0total = 0for batch_idx, (data, target) in enumerate(train_loader):data, target = data.to(device), target.to(device)  # 移至GPU(如果可用)optimizer.zero_grad()  # 梯度清零output = model(data)  # 前向傳播loss = criterion(output, target)  # 計算損失loss.backward()  # 反向傳播optimizer.step()  # 更新參數# 記錄當前 iteration 的損失（注意：這里直接使用單 batch 損失，而非累加平均）iter_loss = loss.item()all_iter_losses.append(iter_loss)iter_indices.append(epoch * len(train_loader) + batch_idx + 1)  # iteration 序號從1開始# 統計準確率和損失（原邏輯保留，用于 epoch 級統計）running_loss += iter_loss_, predicted = output.max(1)total += target.size(0)correct += predicted.eq(target).sum().item()# 每100個批次打印一次訓練信息（可選：同時打印單 batch 損失）if (batch_idx + 1) % 100 == 0:print(f'Epoch: {epoch+1}/{epochs} | Batch: {batch_idx+1}/{len(train_loader)} 'f'| 單Batch損失: {iter_loss:.4f} | 累計平均損失: {running_loss/(batch_idx+1):.4f}')# 原 epoch 級邏輯（測試、打印 epoch 結果）不變epoch_train_loss = running_loss / len(train_loader)epoch_train_acc = 100. * correct / totalepoch_test_loss, epoch_test_acc = test(model, test_loader, criterion, device)print(f'Epoch {epoch+1}/{epochs} 完成 | 訓練準確率: {epoch_train_acc:.2f}% | 測試準確率: {epoch_test_acc:.2f}%')# 繪制所有 iteration 的損失曲線plot_iter_losses(all_iter_losses, iter_indices)# 保留原 epoch 級曲線（可選）# plot_metrics(train_losses, test_losses, train_accuracies, test_accuracies, epochs)return epoch_test_acc  # 返回最終測試準確率# 6. 測試模型
def test(model, test_loader, criterion, device):model.eval()  # 設置為評估模式test_loss = 0correct = 0total = 0with torch.no_grad():  # 不計算梯度，節省內存和計算資源for data, target in test_loader:data, target = data.to(device), target.to(device)output = model(data)test_loss += criterion(output, target).item()_, predicted = output.max(1)total += target.size(0)correct += predicted.eq(target).sum().item()avg_loss = test_loss / len(test_loader)accuracy = 100. * correct / totalreturn avg_loss, accuracy  # 返回損失和準確率# 7.繪制每個 iteration 的損失曲線
def plot_iter_losses(losses, indices):plt.figure(figsize=(10, 4))plt.plot(indices, losses, 'b-', alpha=0.7, label='Iteration Loss')plt.xlabel('Iteration（Batch序號）')plt.ylabel('損失值')plt.title('每個 Iteration 的訓練損失')plt.legend()plt.grid(True)plt.tight_layout()plt.show()# 8. 執行訓練和測試（設置 epochs=2 驗證效果）
epochs = 2  
print("開始訓練模型...")
final_accuracy = train(model, train_loader, test_loader, criterion, optimizer, device, epochs)
print(f"訓練完成！最終測試準確率: {final_accuracy:.2f}%")