diff --git a/develop/doc/_sources/getstarted/build_and_install/docker_install_en.rst.txt b/develop/doc/_sources/getstarted/build_and_install/docker_install_en.rst.txt
index b6a9dd963c603ebd2950e94c65d3e0c64a410b44..03df497506099d2fb758bd0ab437d2c082f2b537 100644
--- a/develop/doc/_sources/getstarted/build_and_install/docker_install_en.rst.txt
+++ b/develop/doc/_sources/getstarted/build_and_install/docker_install_en.rst.txt
@@ -210,7 +210,7 @@ compiling tools and dependencies.
 
    .. code-block:: bash
 
-      docker build -t paddle:prod -f build/Dockerfile .
+      docker build -t paddle:prod -f build/Dockerfile ./build
 
    The above command will generate the production image by copying the compiled PaddlePaddle program into the image.
 
diff --git a/develop/doc/getstarted/build_and_install/docker_install_en.html b/develop/doc/getstarted/build_and_install/docker_install_en.html
index 919ebf239af895442ee3eef7aedb494fe5f24662..e4c0bbdfae12d445e182006c5d49ba6db7219bca 100644
--- a/develop/doc/getstarted/build_and_install/docker_install_en.html
+++ b/develop/doc/getstarted/build_and_install/docker_install_en.html
@@ -381,7 +381,7 @@ docker build -t paddle:dev .
 </div>
 <p>The above command will compile PaddlePaddle and create a Dockerfile for building production image. All the generated files are in the build directory. &#8220;WITH_GPU&#8221; controls if the generated production image supports GPU. &#8220;WITH_AVX&#8221; controls if the generated production image supports AVX. &#8220;WITH_TEST&#8221; controls if the unit test will be generated.</p>
 <p>The second step is to run:</p>
-<div class="highlight-bash"><div class="highlight"><pre><span></span>docker build -t paddle:prod -f build/Dockerfile .
+<div class="highlight-bash"><div class="highlight"><pre><span></span>docker build -t paddle:prod -f build/Dockerfile ./build
 </pre></div>
 </div>
 <p>The above command will generate the production image by copying the compiled PaddlePaddle program into the image.</p>
diff --git a/develop/doc_cn/_sources/getstarted/build_and_install/docker_install_cn.rst.txt b/develop/doc_cn/_sources/getstarted/build_and_install/docker_install_cn.rst.txt
index e35dc6aaef090dc22a6558515c281c9c8fc821cb..da2d4234658b6ea4730346e721437cc1633c4362 100644
--- a/develop/doc_cn/_sources/getstarted/build_and_install/docker_install_cn.rst.txt
+++ b/develop/doc_cn/_sources/getstarted/build_and_install/docker_install_cn.rst.txt
@@ -155,7 +155,7 @@ PaddlePaddle Book是为用户和开发者制作的一个交互式的Jupyter Node
 
    .. code-block:: bash
       
-      docker build -t paddle:prod -f build/Dockerfile .
+      docker build -t paddle:prod -f build/Dockerfile ./build
 
    以上命令会按照生成的Dockerfile把生成的程序拷贝到生产镜像中并做相应的配置，最终生成名为paddle:prod的生产镜像。
 
diff --git a/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html b/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
index ab4f7817f9de17cd3df97fee3624f68f89661070..47e3b4a722b087db8bd8679afdef1b231fe989fd 100644
--- a/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
+++ b/develop/doc_cn/getstarted/build_and_install/docker_install_cn.html
@@ -335,7 +335,7 @@ docker build -t paddle:dev .
 </div>
 <p>以上命令会编译PaddlePaddle，生成运行程序，以及生成创建生产镜像的Dockerfile。所有生成的的文件都在build目录下。“WITH_GPU”控制生成的生产镜像是否支持GPU，“WITH_AVX”控制生成的生产镜像是否支持AVX，”WITH_TEST“控制是否生成单元测试。</p>
 <p>第二步是运行：</p>
-<div class="highlight-bash"><div class="highlight"><pre><span></span>docker build -t paddle:prod -f build/Dockerfile .
+<div class="highlight-bash"><div class="highlight"><pre><span></span>docker build -t paddle:prod -f build/Dockerfile ./build
 </pre></div>
 </div>
 <p>以上命令会按照生成的Dockerfile把生成的程序拷贝到生产镜像中并做相应的配置，最终生成名为paddle:prod的生产镜像。</p>