diff --git a/examples/aerodynamic_car_design/README.md b/examples/aerodynamic_car_design/README.md
index 883c9b8..68737fd 100755
--- a/examples/aerodynamic_car_design/README.md
+++ b/examples/aerodynamic_car_design/README.md
@@ -22,7 +22,7 @@ volume | 0.0211 | 0.0253
 
 
 # 3. Enviroment
-## datadownload
+## 3.1 Linux: data and checkpoint download
 ``` Data
 cd examples/aerodynamic_car_design/
 mkdir -p ./data && cd ./data
@@ -37,6 +37,11 @@ wget https://paddle-org.bj.bcebos.com/paddlecfd/checkpoints/pptransformer/model_
 cd .. && cd ..
 ```
 
+## 3.2 Windows: data and checkpoint download
+
+[click_me_to_download_data](https://paddle-org.bj.bcebos.com/paddlecfd/datasets/pptransformer/mlcfd_data.zip)
+
+[click_me_to_download_checkpoint](https://paddle-org.bj.bcebos.com/paddlecfd/checkpoints/pptransformer/model_131.pdparams)
 
 # 4. How to Run
 ## ShapeNet-Car
@@ -51,6 +56,10 @@ python main_shapenetcar.py \
     checkpoint=./checkpoint/shapenet_car/model_131.pdparams
 ```
 
+if test successfully:
+<img width="1825" height="707" alt="image" src="https://github.com/user-attachments/assets/0e44b21a-e2ff-440e-9663-f0848d88a274" />
+
+
 ## DrivAerNet++ (TODO)
 ```sh
 python -m paddle.distributed.launch --gpus=0,1 main_v2.py \
diff --git a/examples/aerodynamic_car_design/main_shapenetcar.py b/examples/aerodynamic_car_design/main_shapenetcar.py
index 5963e51..e5dfaf7 100755
--- a/examples/aerodynamic_car_design/main_shapenetcar.py
+++ b/examples/aerodynamic_car_design/main_shapenetcar.py
@@ -24,8 +24,7 @@
 import paddle
 from paddle.io import DataLoader
 from tqdm import tqdm
-from transolver import Model
-
+from ppcfd.models import Transolver as Model
 from ppcfd.data.shapenetcar_datamodule import GraphDataset
 from ppcfd.data.shapenetcar_datamodule import load_train_val_fold
 
@@ -264,8 +263,8 @@ def train(
     np.savetxt(f"{path}/val_loss_{config.num_epochs}.txt", val_loss_list)
     end = time.time()
     time_elapsed = end - start
-    log.info("Number of parameters:", get_nb_trainable_params(model))
-    log.info("Time elapsed: {0:.2f} seconds".format(time_elapsed))
+    log.info(f"Number of parameters: {get_nb_trainable_params(model)}")
+    log.info(f"Time elapsed: {time_elapsed} seconds")
     return model