diff --git a/fluid/faster_rcnn/README.md b/fluid/faster_rcnn/README.md
index 617a103278e29c22e5aef73830605522159b4a86..b676cce670989e3e5f1f0b7014cebe96cd27fc42 100644
--- a/fluid/faster_rcnn/README.md
+++ b/fluid/faster_rcnn/README.md
@@ -57,6 +57,7 @@ After data preparation, one can start the training step by:
     sh ./pretrained/download.sh
 
 Set `pretrained_model` to load pre-trained model. In addition, this parameter is used to load trained model when finetuning as well.
+Please make sure that pretrained_model is downloaded and loaded correctly, otherwise, the loss may be NAN during training.
 
 **Install the [cocoapi](https://github.com/cocodataset/cocoapi):**
 
diff --git a/fluid/faster_rcnn/README_cn.md b/fluid/faster_rcnn/README_cn.md
index 6f9e892e6b64008fdf23bc53466d504d04266fb7..b17386a9e07139d0ff1bec58fe560609f7fac7bc 100644
--- a/fluid/faster_rcnn/README_cn.md
+++ b/fluid/faster_rcnn/README_cn.md
@@ -57,6 +57,7 @@ Faster RCNN 目标检测模型
     sh ./pretrained/download.sh
 
 通过初始化`pretrained_model` 加载预训练模型。同时在参数微调时也采用该设置加载已训练模型。
+请在训练前确认预训练模型下载与加载正确，否则训练过程中损失可能会出现NAN。
 
 **安装[cocoapi](https://github.com/cocodataset/cocoapi)：**