diff --git a/deploy/cpp/CMakeLists.txt b/deploy/cpp/CMakeLists.txt
index 349afa2cae5bf40721cafdf38bbf28ddd621beeb..a54979683cd14d2a352cb789b9d6dc7bd26d0a46 100644
--- a/deploy/cpp/CMakeLists.txt
+++ b/deploy/cpp/CMakeLists.txt
@@ -320,46 +320,34 @@ target_link_libraries(video_segmenter ${DEPS})
 
 if (WIN32 AND WITH_MKL)
     add_custom_command(TARGET classifier POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/mkldnn.dll
     )
     add_custom_command(TARGET detector POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/mkldnn.dll
     )
     add_custom_command(TARGET segmenter POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/mkldnn.dll
     )
     add_custom_command(TARGET video_classifier POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/mkldnn.dll
     )
     add_custom_command(TARGET video_detector POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/kldnn.dll
     )
     add_custom_command(TARGET video_segmenter POST_BUILD
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/Release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/Release/libiomp5md.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/Release/mkldnn.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./release/mklml.dll
-        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./release/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/mklml.dll ./paddlex_inference/mklml.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mklml/lib/libiomp5md.dll ./paddlex_inference/libiomp5md.dll
+        COMMAND ${CMAKE_COMMAND} -E copy_if_different ${PADDLE_DIR}/third_party/install/mkldnn/lib/mkldnn.dll ./paddlex_inference/mkldnn.dll
     )
     # for encryption
     if (EXISTS "${ENCRYPTION_DIR}/lib/pmodel-decrypt.dll")
diff --git a/docs/deploy/server/cpp/linux.md b/docs/deploy/server/cpp/linux.md
index 452c49bec8d4f60a6288fae40824e416729246b2..acfd1477a5d8ea09d4db3301902f89041dd962fe 100644
--- a/docs/deploy/server/cpp/linux.md
+++ b/docs/deploy/server/cpp/linux.md
@@ -41,7 +41,7 @@ fluid_inference
 └── version.txt # 版本和编译信息
 ```
 
-**注意:** 预编译版本除nv-jetson-cuda10-cudnn7.5-trt5 以外其它包都是基于`GCC 4.8.5`编译，使用高版本`GCC`可能存在 `ABI`兼容性问题，建议降级或[自行编译预测库](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/advanced_guide/inference_deployment/inference/build_and_install_lib_cn.html#id12)。
+**注意:** 预编译版本除`nv-jetson-cuda10-cudnn7.5-trt5` 以外其它包都是基于`GCC 4.8.5`编译，使用高版本`GCC`可能存在 `ABI`兼容性问题，建议降级或[自行编译预测库](https://www.paddlepaddle.org.cn/documentation/docs/zh/develop/advanced_guide/inference_deployment/inference/build_and_install_lib_cn.html#id12)。
 
 
 ### Step3: 编译
@@ -97,7 +97,7 @@ make
 **注意：** linux环境下编译会自动下载OPENCV, PaddleX-Encryption和YAML，如果编译环境无法访问外网，可手动下载：
 
 - [opencv3.4.6gcc4.8ffmpeg.tar.gz2](https://bj.bcebos.com/paddleseg/deploy/opencv3.4.6gcc4.8ffmpeg.tar.gz2)
-- [paddlex-encryption.zip](https://bj.bcebos.com/paddlex/tools/paddlex-encryption.zip)
+- [paddlex-encryption.zip](https://bj.bcebos.com/paddlex/tools/1.2.0/paddlex-encryption.zip)
 - [yaml-cpp.zip](https://bj.bcebos.com/paddlex/deploy/deps/yaml-cpp.zip)
 
 opencv3gcc4.8.tar.bz2文件下载后解压，然后在script/build.sh中指定`OPENCE_DIR`为解压后的路径。
diff --git a/docs/deploy/server/cpp/windows.md b/docs/deploy/server/cpp/windows.md
index c39d96247b099865f7aba1140df79351141acb4c..7aecc3462299e81f3a4b56151619abef04271522 100644
--- a/docs/deploy/server/cpp/windows.md
+++ b/docs/deploy/server/cpp/windows.md
@@ -82,7 +82,7 @@ PaddlePaddle C++ 预测库针对是否使用GPU、是否支持TensorRT、以及
 1. 如果使用`CPU`版预测库，请把`WITH_GPU`的`值`去掉勾
 2. 如果使用的是`openblas`版本，请把`WITH_MKL`的`值`去掉勾
 3. Windows环境下编译会自动下载YAML，如果编译环境无法访问外网，可手动下载： [yaml-cpp.zip](https://bj.bcebos.com/paddlex/deploy/deps/yaml-cpp.zip)。YAML文件下载后无需解压，在`cmake/yaml.cmake`中将`URL https://bj.bcebos.com/paddlex/deploy/deps/yaml-cpp.zip` 中的网址，改为下载文件的路径。
-4. 如果需要使用模型加密功能，需要手动下载[Windows预测模型加密工具](https://bj.bcebos.com/paddlex/tools/win/paddlex-encryption.zip)。例如解压到`D:/projects`，解压后目录为`D:/projects/paddlex-encryption`。编译时需勾选`WITH_EBNCRYPTION`并且在`ENCRTYPTION_DIR`填入`D:/projects/paddlex-encryption`。
+4. 如果需要使用模型加密功能，需要手动下载[Windows预测模型加密工具](https://bj.bcebos.com/paddlex/tools/win/1.2.0/paddlex-encryption.zip)。例如解压到`D:/projects`，解压后目录为`D:/projects/paddlex-encryption`。编译时需勾选`WITH_EBNCRYPTION`并且在`ENCRTYPTION_DIR`填入`D:/projects/paddlex-encryption`。
 ![](../../images/vs2019_step_encryption.png)
 ![](../../images/vs2019_step6.png)
 **设置完成后**, 点击上图中`保存并生成CMake缓存以加载变量`。
diff --git a/docs/deploy/server/encryption.md b/docs/deploy/server/encryption.md
index c172cc802bc859f427e13f5684f092a5b8c5fc1f..1c7c132606dd3cfbff3a8da69c4e5a797516a59f 100644
--- a/docs/deploy/server/encryption.md
+++ b/docs/deploy/server/encryption.md
@@ -40,9 +40,9 @@ PaddleX提供一个轻量级的模型加密部署方案，通过PaddleX内置的
 
 ### 1.2 加密工具
 
-[Linux版本 PaddleX模型加密工具](https://bj.bcebos.com/paddlex/tools/paddlex-encryption.zip)，编译脚本会自动下载该版本加密工具，您也可以选择手动下载。
+[Linux版本 PaddleX模型加密工具](https://bj.bcebos.com/paddlex/tools/1.2.0/paddlex-encryption.zip)，编译脚本会自动下载该版本加密工具，您也可以选择手动下载。
 
-[Windows版本 PaddleX模型加密工具](https://bj.bcebos.com/paddlex/tools/win/paddlex-encryption.zip)，该版本加密工具需手动下载，如果您在使用Visual Studio 2019编译C++预测代码的过程中已经下载过该工具，此处可不必重复下载。
+[Windows版本 PaddleX模型加密工具](https://bj.bcebos.com/paddlex/tools/win/1.2.0/paddlex-encryption.zip)，该版本加密工具需手动下载，如果您在使用Visual Studio 2019编译C++预测代码的过程中已经下载过该工具，此处可不必重复下载。
 
 Linux加密工具包含内容为：
 ```