From cae9640ce0e6350b404dc956128b209d906b1db6 Mon Sep 17 00:00:00 2001 From: tink2123 Date: Wed, 31 Oct 2018 13:40:10 +0800 Subject: [PATCH 1/2] update windows_cpp_inference.md --- .../howto/inference/windows_cpp_inference.md | 47 ++++++++++++------- 1 file changed, 31 insertions(+), 16 deletions(-) diff --git a/doc/fluid/user_guides/howto/inference/windows_cpp_inference.md b/doc/fluid/user_guides/howto/inference/windows_cpp_inference.md index f16c95111c0..edb628e6f05 100644 --- a/doc/fluid/user_guides/howto/inference/windows_cpp_inference.md +++ b/doc/fluid/user_guides/howto/inference/windows_cpp_inference.md @@ -6,7 +6,7 @@ Windows环境模型预测使用说明 ### 硬件环境 -测试环境硬件配置: +测试环境硬件配置: | CPU | I7-8700K | |----------|:-------------:| @@ -28,14 +28,15 @@ Windows环境模型预测使用说明 需要去NVIDIA官网[https://www.geforce.cn/drivers](https://www.geforce.cn/drivers) 下载显卡对应的驱动。推荐391版本 - -![](image/image1.png) - +

+ +

安装时需要勾选自定义,勾选安装全部。 验证安装需要进入cmd中,输入nvcc -V查看。 - -![](image/image2.png) +

+ +

如果有显卡安装驱动,也可以选择直接安装CUDA8.0,[https://developer.nvidia.com/cuda-80-ga2-download-archive](https://developer.nvidia.com/cuda-80-ga2-download-archive) @@ -75,37 +76,51 @@ Cmake可以在官网进行下载,并添加到环境变量中。[[https://cmake 执行完毕后,build目录如图所示,打开 箭头指向的solution文件: -![](image/image3.png) +

+ +

修改编译属性为/MT: -![](image/image4.png) +

+ +

-![](image/image5.png) +

+ +

编译生成选项改成Release。 -![](image/image6.png) +

+ +

-![](image/image7.png) +

+ +

将提供的依赖包中,Release下的openblas和模型文件拷贝到编译生成的Release下。 -![](image/image8.png) +

+ +

通过cmd进到Release目录执行: 1. 开启GLOG - set GLOG\_v=3 + set GLOG\_v=3 2. 进行预测 - simple\_on\_word2vec.exe \--dirname=.\\word2vec.inference.model + simple\_on\_word2vec.exe \--dirname=.\\word2vec.inference.model -![](image/image9.png) +

+ +

-FAQ: +**FAQ:** 路径中尽量不要包含空格,例如发现CUDA\_LIB路径是Program Files(x86)可能会出错。可以将CUDA拷贝到一个新位置(这里直接拷贝就行) From 4131dba8b4341823006bfd06fd150984f24397f8 Mon Sep 17 00:00:00 2001 From: tink2123 Date: Wed, 31 Oct 2018 13:41:08 +0800 Subject: [PATCH 2/2] update index --- doc/fluid/user_guides/howto/inference/index.rst | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/doc/fluid/user_guides/howto/inference/index.rst b/doc/fluid/user_guides/howto/inference/index.rst index 3d52e071ff3..1e572b2d4df 100644 --- a/doc/fluid/user_guides/howto/inference/index.rst +++ b/doc/fluid/user_guides/howto/inference/index.rst @@ -10,4 +10,4 @@ PaddlePaddle Fluid 提供了 C++ API 来支持模型的部署上线 build_and_install_lib_cn.rst native_infer.md paddle_tensorrt_infer.md - windows_inference.md + windows_cpp_inference.md