diff --git a/docs/zh_CN/tutorials/quick_start.md b/docs/zh_CN/tutorials/quick_start.md
index 0a021185f6b58c4af1a8aa1c3de946cce6861736..5a70de4921b20c1b11a774f54c0a3f77cb989009 100644
--- a/docs/zh_CN/tutorials/quick_start.md
+++ b/docs/zh_CN/tutorials/quick_start.md
@@ -73,9 +73,9 @@ python tools/download.py -a MobileNetV3_large_x1_0 -p ./pretrained -d True
 
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/ResNet50_vd.yaml
 
 ```
@@ -91,9 +91,9 @@ python -m paddle.distributed.launch \  
 
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/ResNet50_vd_finetune.yaml
 
 ```
@@ -119,9 +119,9 @@ pretrained_model: "./pretrained/ResNet50_vd_ssld_pretrained"
 训练脚本如下。
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/ResNet50_vd_ssld_finetune.yaml
 ```
 
@@ -134,9 +134,9 @@ python -m paddle.distributed.launch \  
 
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/MobileNetV3_large_x1_0_finetune.yaml
 ```
 
@@ -150,9 +150,9 @@ python -m paddle.distributed.launch \  
 
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/ResNet50_vd_ssld_random_erasing_finetune.yaml
 ```
 
@@ -189,9 +189,9 @@ TRAIN:
 
 ```shell
 export CUDA_VISIBLE_DEVICES=0
-python -m paddle.distributed.launch \  
-    --selected_gpus="0" \  
-    tools/train.py \  
+python -m paddle.distributed.launch \
+    --selected_gpus="0" \
+    tools/train.py \
         -c ./configs/quick_start/R50_vd_distill_MV3_large_x1_0.yaml
 ```