diff --git a/README.md b/README.md
index b1cc15d261a396ba8a8846375854283324a5f6a0..a8f9d6d67dc34855a59aea24f8d54f2fda92fd95 100644
--- a/README.md
+++ b/README.md
@@ -47,7 +47,7 @@ We adopt yolox as the detection method. We used stronger data augmentation and r
 
 ### Training 
 
-Download the pretrained model at [Onedrive](https://mailstsinghuaeducn-my.sharepoint.com/:f:/g/personal/liuxy21_mails_tsinghua_edu_cn/EkCTrRfHUZVEtD7eHwLkYSkBCTXlh9ekDteSzK6jM4oo-A?e=m0aNCy) (password: groupji) and put it in the folder `pretrained_models/yolox`. Then use the following command:
+Download the pretrained model at [Onedrive](https://mailstsinghuaeducn-my.sharepoint.com/:f:/g/personal/liuxy21_mails_tsinghua_edu_cn/EkCTrRfHUZVEtD7eHwLkYSkBCTXlh9ekDteSzK6jM4oo-A?e=m0aNCy) (password: groupji)/[BaiDuYunPan](链接: https://pan.baidu.com/s/1AU7DGCmZWsH9VgQnbTRjow  密码: aw68) and put it in the folder `pretrained_models/yolox`. Then use the following command:
 
 `./det/yolox/tools/train_yolox.sh <config_path> <gpu_ids> (other args)`