Skip to content

1031 windows infer #256

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 2 commits into from
Oct 31, 2018
Merged

1031 windows infer #256

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
cae9640
update windows_cpp_inference.md
tink2123 Oct 31, 2018
4131dba
update index
tink2123 Oct 31, 2018
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion doc/fluid/user_guides/howto/inference/index.rst
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -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
47 changes: 31 additions & 16 deletions doc/fluid/user_guides/howto/inference/windows_cpp_inference.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -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 |
|----------|:-------------:|
Expand All @@ -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)

<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image1.png" >
</p>
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)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image2.png">
</p>

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)

Expand Down Expand Up @@ -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)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image3.png">
</p>

修改编译属性为/MT：

![](image/image4.png)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image4.png">
</p>

![](image/image5.png)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image5.png">
</p>

编译生成选项改成Release。

![](image/image6.png)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image6.png">
</p>

![](image/image7.png)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image7.png">
</p>

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)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image8.png">
</p>

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)
<p align="center">
<img src="https://raw.githubusercontent.com/PaddlePaddle/FluidDoc/develop/doc/fluid/user_guides/howto/inference/image/image9.png">
</p>

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）