fix bug

Image/ResNet/code/train.py

@@ -4,14 +4,14 @@ sys.path.append(os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(
 from utils.dataset_utils import get_cifar10_dataloaders
 from utils.train_utils import train_model, train_model_data_augmentation, train_model_backdoor
 from utils.parse_args import parse_args
-from model import ResNet
+from model import ResNet34
 
 def main():
     # 解析命令行参数
     args = parse_args()
     
     # 创建模型
-    model = ResNet()
+    model = ResNet34()
     
     if args.train_type == '0':
         # 获取数据加载器

Image/utils/train_utils.py

@@ -189,8 +189,8 @@ def train_model(model, trainloader, testloader, epochs=200, lr=0.1, device='cuda
         logger.info(f'Epoch: {epoch+1} | Test Loss: {test_loss/(batch_idx+1):.3f} | '
                    f'Test Acc: {acc:.2f}%')
         
-        # 每1个epoch保存一次
-        if (epoch + 1) % 1 == 0:
+        # 每5个epoch保存一次
+        if (epoch + 1) % 5 == 0:
             # 创建一个专门用于收集embedding的顺序dataloader
             ordered_loader = torch.utils.data.DataLoader(
                 trainloader.dataset,  # 使用相同的数据集
@@ -198,7 +198,7 @@ def train_model(model, trainloader, testloader, epochs=200, lr=0.1, device='cuda
                 shuffle=False,  # 确保顺序加载
                 num_workers=trainloader.num_workers
             )
-            save_model = time_travel_saver(model, ordered_loader, device, save_dir, model_name, interval = 1)
+            save_model = time_travel_saver(model, ordered_loader, device, save_dir, model_name, interval = 1, auto_save_embedding = True)
             save_model.save()
             
         scheduler.step()

ttv_utils/__init__.py

@@ -52,6 +52,6 @@
 """
 
 from .feature_predictor import FeaturePredictor, predict_feature
-from .save_embeddings import time_travel_savers
+from .save_embeddings import time_travel_saver
 
 __all__ = ['FeaturePredictor', 'predict_feature', 'time_travel_saver']

ttv_utils/save_embeddings.py

@@ -15,7 +15,7 @@ class time_travel_saver:
     4. 标签数据 (label/labels.npy)
     """
     
-    def __init__(self, model, dataloader, device, save_dir, model_name, interval=1):
+    def __init__(self, model, dataloader, device, save_dir, model_name, interval=1, auto_save_embedding=False):
         """初始化
         
         Args:
@@ -32,6 +32,7 @@ class time_travel_saver:
         self.save_dir = save_dir
         self.model_name = model_name
         self.interval = interval
+        self.auto_save = auto_save_embedding
         
         # 创建保存目录结构
         self.model_dir = os.path.join(save_dir, 'model')
@@ -160,17 +161,19 @@ class time_travel_saver:
         # 保存模型权重
         model_path = os.path.join(self.model_dir, f'{self.current_epoch}.pth')
         torch.save(self.model.state_dict(), model_path)
+        
+        if self.auto_save:
+            # 提取并保存特征和预测结果
+            features, predictions = self._extract_features_and_predictions()
             
-        # 提取并保存特征和预测结果
-        features, predictions = self._extract_features_and_predictions()
-            
-        # 保存特征
-        np.save(os.path.join(self.repr_dir, f'{self.current_epoch}.npy'), features)
+            # 保存特征
+            np.save(os.path.join(self.repr_dir, f'{self.current_epoch}.npy'), features)
             
-        # 保存预测结果
-        np.save(os.path.join(self.pred_dir, f'{self.current_epoch}.npy'), predictions)
+            # 保存预测结果
+            np.save(os.path.join(self.pred_dir, f'{self.current_epoch}.npy'), predictions)
             
-        print(f"Epoch {self.current_epoch * self.interval} 的数据已保存:")
-        print(f"- 模型权重: {model_path}")
-        print(f"- 特征向量: [样本数: {features.shape[0]}, 特征维度: {features.shape[1]}]")
-        print(f"- 预测结果: [样本数: {predictions.shape[0]}, 类别数: {predictions.shape[1]}]")
+            print(f"Epoch {self.current_epoch * self.interval} 的数据已保存:")
+            print(f"- 模型权重: {model_path}")
+            print(f"- 特征向量: [样本数: {features.shape[0]}, 特征维度: {features.shape[1]}]")
+            print(f"- 预测结果: [样本数: {predictions.shape[0]}, 类别数: {predictions.shape[1]}]")
+        print(f"Epoch {self.current_epoch * self.interval} 的数据已保存")