diff --git a/model_examples/MapTR/README.md b/model_examples/MapTR/README.md
index 505f654bb71dc40836e05dedb3adaec518d9d52e..761bc48370b9498c76a575fab91f2ecf7cede152 100644
--- a/model_examples/MapTR/README.md
+++ b/model_examples/MapTR/README.md
@@ -207,7 +207,7 @@ MapTR是一种高效的端到端Transformer模型，用于在线构建矢量化
   1） 创建编译优化基础镜像。以arm的镜像为例：
 
   - arm镜像地址：
-      https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/docker_images/pytorcharm_compile.tar.gz
+      https://pytorch-package.obs.cn-north-4.myhuaweicloud.com/images/pytorcharm_compile.tar.gz
 
     将镜像的image_id记为image_id，将创建容器时的宿主机路径和容器路径分别记为path1和path2：
     ```
diff --git a/tests/README.md b/tests/README.md
index aae82d60c804f93d09c827b10d5e4c4af463ebc7..5607e9f154b28de412db6a9110be4d0e9d36b6d7 100644
--- a/tests/README.md
+++ b/tests/README.md
@@ -17,6 +17,7 @@
     ```Bash
     git clone -b 1.x https://github.com/open-mmlab/mmcv.git
     cd mmcv
+    MMCV_WITH_OP=1 FORCE_NPU=1 python setup.
     MMCV_WITH_OP=1 FORCE_NPU=1 python setup.py install
     cd ../
     ```
diff --git a/tests/requirements.txt b/tests/requirements.txt
index 4873e1df52f430840336eec0580feaadb51258ed..6ba7541902489a4d869e3bf5296a6f595a283f0e 100644
--- a/tests/requirements.txt
+++ b/tests/requirements.txt
@@ -1,3 +1,5 @@
 torch_scatter==2.1.2
 torchvision==0.16.0
-hypothesis
\ No newline at end of file
+hypothesis
+expecttest
+pyyaml
\ No newline at end of file