增加成三分类。

train/food_classifier.py

@@ -18,7 +18,7 @@ plt.rcParams['axes.unicode_minus'] = False  # 解决保存图像是负号'-'显
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 print(f"使用设备: {device}")
 
-# 定义CNN模型（基于CIFAR10结构，输出层改为2分类）
+# 定义CNN模型（基于CIFAR10结构，输出层改为3分类）
 class FoodCNN(nn.Module):
     def __init__(self):
         super(FoodCNN, self).__init__()
@@ -43,7 +43,7 @@ class FoodCNN(nn.Module):
         # 全连接层
         self.fc1 = nn.Linear(128 * 4 * 4, 512)
         self.dropout4 = nn.Dropout(0.5)
-        self.fc2 = nn.Linear(512, 2)  # 改为2分类
+        self.fc2 = nn.Linear(512, 3)  # 改为2分类
         
     def forward(self, x):
         # 第一个卷积块
@@ -150,8 +150,8 @@ def test(model, test_loader, device, class_names):
     model.eval()
     correct = 0
     total = 0
-    class_correct = list(0. for i in range(2))
-    class_total = list(0. for i in range(2))
+    class_correct = list(0. for i in range(3))
+    class_total = list(0. for i in range(3))
     
     with torch.no_grad():
         test_bar = tqdm(test_loader, desc='测试中')
@@ -174,7 +174,7 @@ def test(model, test_loader, device, class_names):
             })
     
     print(f'\n测试集总体准确率: {100.*correct/total:.2f}%')
-    for i in range(2):
+    for i in range(3):
         if class_total[i] > 0:
             print(f'{class_names[i]} 准确率: {100.*class_correct[i]/class_total[i]:.2f}%')
     
@@ -216,7 +216,7 @@ if __name__ == '__main__':
     val_accuracies = []
 
     best_val_acc = 0.0
-    best_model_path = '../model/01/best_food_model.pth'
+    best_model_path = '../model/02/best_food_model.pth'
 
     print("开始训练...")
     for epoch in range(num_epochs):