diff --git a/baseline/compare_models.py b/baseline/compare_models.py
index 7175071..b0e3c4c 100644
--- a/baseline/compare_models.py
+++ b/baseline/compare_models.py
@@ -5,7 +5,7 @@ baseline/compare_models.py
 author: yukun-hh
 date: 2026-5-14
 """
-import sys, os
+import sys, os, re
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 import numpy as np
@@ -29,7 +29,7 @@ matplotlib.rcParams['axes.unicode_minus'] = False
 # ============================================================
 # ★★★ 可配置参数 ★★★
 # ============================================================
-DATA_ROOT   = '../trash_division_data/ultimate_4_class/'
+DATA_ROOT   = '../../trash_division_data/ultimate_4_class/'
 BATCH_SIZE  = 32
 IMAGE_SIZE  = 256
 NUM_WORKERS = 4
@@ -45,7 +45,7 @@ NUM_CLASSES = 4
 
 def get_resnet34_preds(train_loader, val_loader, device):
     model = Net(num_classes=NUM_CLASSES)
-    state_dict = torch.load('best_model.pth', map_location='cpu')
+    state_dict = torch.load('../best_model.pth', map_location='cpu')
     if 'model_state_dict' in state_dict:
         state_dict = state_dict['model_state_dict']
     elif 'model' in state_dict:
@@ -106,6 +106,29 @@ def compute_macro_roc(y_true, y_probs):
     return all_fpr, mean_tpr, macro_auc
 
 
+def sanitize_filename(name):
+    return re.sub(r'[^\w\-_]', '_', name).strip('_')
+
+
+def preds_csv_path(out_dir, model_name):
+    safe = sanitize_filename(model_name)
+    return os.path.join(out_dir, f'{safe}_preds.csv')
+
+
+def save_preds_csv(path, y_true, y_preds, y_probs):
+    header = 'y_true,y_pred,' + ','.join(f'prob_{c}' for c in range(NUM_CLASSES))
+    data = np.column_stack([y_true.astype(float), y_preds.astype(float), y_probs])
+    np.savetxt(path, data, delimiter=',', header=header, comments='', fmt='%.6f')
+
+
+def load_preds_csv(path):
+    data = np.loadtxt(path, delimiter=',', skiprows=1)
+    y_true = data[:, 0].astype(int)
+    y_preds = data[:, 1].astype(int)
+    y_probs = data[:, 2:2 + NUM_CLASSES]
+    return y_true, y_preds, y_probs
+
+
 if __name__ == '__main__':
     out_dir = os.path.dirname(os.path.abspath(__file__))
 
@@ -132,12 +155,19 @@ if __name__ == '__main__':
     val_loader = DataLoader(val_dataset, batch_size=BATCH_SIZE, shuffle=False,
                             num_workers=NUM_WORKERS, pin_memory=True, drop_last=False)
 
-    # ———— 评估所有模型 ————
+    # ———— 评估所有模型（有缓存则跳过）————
     results = {}
     for name, func in MODELS:
         print(f"\n{'='*50}")
-        print(f"评估: {name}")
-        y_true, y_preds, y_probs = func(train_loader, val_loader, device)
+        csv_path = preds_csv_path(out_dir, name)
+        if os.path.exists(csv_path):
+            print(f"加载缓存: {os.path.basename(csv_path)}")
+            y_true, y_preds, y_probs = load_preds_csv(csv_path)
+        else:
+            print(f"评估: {name}")
+            y_true, y_preds, y_probs = func(train_loader, val_loader, device)
+            save_preds_csv(csv_path, y_true, y_preds, y_probs)
+            print(f"  预测数据已保存: {os.path.basename(csv_path)}")
         acc = accuracy_score(y_true, y_preds)
         fpr, tpr, roc_auc = compute_macro_roc(y_true, y_probs)
         results[name] = {'y_true': y_true, 'y_preds': y_preds, 'y_probs': y_probs,