Request for All Windows 10 installation package file

[leekwunfung817]

I want all your installation file, that you are build success.
I must make sure that I can build it successfully.

I am using the following things:
1. Windows 10
2. GeForce GT 710
3. (OpenCV 3.4.0, cuDNN 7, CUDA 9, Visual Studio 2015)

but there are too many dependencies software version are not match
I have too much bug to solve when build the darknet.sln or others visual studio project
Please, I have been use a month to try to build this project, but the build is always fail.
reinstall any thing and try many times and buy many types of Hardware, but not working.

Can you tell us your software install information about:
1. Graphic Card Full Version
2. NVIDIA Driver Full Version (include 8xx.xxx.xxx.xxx)
3. Visual Studio Full Version (include v14.xx.xx.xx)
4. CUDA Full Version (include xxx.xxx.xxx)
5. cuDNN Full Version (include xxx.xxx.xxx)

Also, I need the exactly version that you are using, and the installation order, then give me the exactly download url.

I must complete the project that the company request me to do.
Please, tell me the detail if you are use it without bugs and problem.
If you can help me, Thank you very much!

[AlexeyAB]

• GPU nVidia GeForce GTX 970 (Maxwell)

• Windows 7 x64

• nVidia driver version (nvidia-smi): 388.19

• MSVS 2015: 14.0.25431.01 Update 3 https://go.microsoft.com/fwlink/?LinkId=532606&clcid=0x409

• OpenCV 3.3.0 https://sourceforge.net/projects/opencvlibrary/files/opencv-win/3.3.0/opencv-3.3.0-vc14.exe/download

• CUDA 9.0 https://developer.nvidia.com/compute/cuda/9.0/Prod/local_installers/cuda_9.0.176_windows-exe

• cuDNN 7.1.3 for CUDA 9.0: https://developer.nvidia.com/compute/machine-learning/cudnn/secure/v7.1.3/prod/9.0_20180414/cudnn-9.0-windows7-x64-v7.1

---

• Can you build darknet_no_gpu.sln successfully?

• What error do you get when you build darknet.sln?

[PeterQuinn925]

I don't know, but googling for the error message msb3721 gives some good hints: https://stackoverflow.com/questions/42600823/msb3721-error-exited-with-code-1

[AlexeyAB]

@leekwunfung817

• Show the fullscreen screenshot of this error
• What errors do prior to VS reporting the MSB3721 error?
• Can you compile any other CUDA applications?
• Can you build darknet_no_gpu.sln successfully?
• Do you try to compile default darknet.sln or did you make your custom sln-file?
• How many MSVS are installed on your computer, and what are installed?

Also read: https://devtalk.nvidia.com/default/topic/989584/how-to-fix-error-msb3721/

[leekwunfung817]

With NVIDIA GPU

• Windows 10
• GPU NVIDIA GeForce GT 710

Install Step (by Order):

1. install Visual Studio 2015 Update 3

• https://download.microsoft.com/download/4/8/f/48f0645f-51b6-4733-b808-63e640cddaec/vs2015.3.exe

2. download OpenCV 3.4.0

• https://nchc.dl.sourceforge.net/project/opencvlibrary/opencv-win/3.4.0/opencv-3.4.0-vc14_vc15.exe

3. install NVIDIA Cuda 9.1

• https://developer.download.nvidia.com/compute/cuda/9.1/secure/Prod/local_installers/cuda_9.1.85_win10.exe?nCAjZE0jUxSE-AcJXMlPk4A3evyLV-gHoreH1tYsooan2BDHNIbM2c4hwYCsVz1DxuPIrVqCqffEipmpGI9EBoXGzuX5C9A05yvvN6S0HBWexrtupkQlXlDGdSGu-5znSfriBqbbPzDWdMBiWc8n-9jl2rC7_6oCI2BYGDX1a1fL-Q

4. download NVIDIA CuDNN 7

• https://developer.download.nvidia.com/compute/machine-learning/cudnn/secure/v7.0.5/prod/9.1_20171129/cudnn-9.1-windows10-x64-v7.zip?gLSaZu2MVANj6BmRhbWbgraHeB4D67aB_223v2QqgdM60BxkRCpWK5wmi2TE8-3FjaTdIxJwabY_wakOSogwnRZIcS5dcnG3T0SZ3GnbKR7Rc_VlLmaU792mgc7d4t6ro0_4OZiVqeysiKOeNLv8_LjC9B7-kwl8sISwXgVpp23qr80UdZ-wpDPhU7txxQhW2sh1A8FFJhdxbYX9

Complie Step:
put CuDNN to NVIDIA GPU Computing Toolkit
setup the darknet.sln properties

• (cudnn-9.1-windows10-x64-v7\cuda\bin\cudnn64_7.dll) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\bin)

• (cudnn-9.1-windows10-x64-v7\cuda\lib\x64\cudnn.lib) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\lib\x64)

• (cudnn-9.1-windows10-x64-v7\cuda\include\cudnn.h) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\include)

1. Set project

• Release
• x64

(Right click darknet project) > (Click Properties)

2. Configuration Properties > C/C++ > Additional include Directories
   add

• D:\opencv_3.4.0\build\include
• D:\opencv_3.4.0\build\include\opencv
• D:\opencv_3.4.0\build\include\opencv2

3. Configuration Properties > Linker > Input > Additional Dependencies
   add

• D:\opencv_3.4.0\build\x64\vc14\lib\opencv_world340.lib
• D:\opencv_3.4.0\build\x64\vc14\lib\opencv_world340d.lib

3. Configuration Properties > CUDA C/C++ > CUDA Toolkit Custom Dir
   add

• C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA

Please use in the MS-DOS command

darknet-master\build\darknet\x64\darknet.exe

1. GPU (darknet.sln)
Finally Build Success
But two warnings

2. No GPU (darknet_no_gpu.sln)
156 Errors 1400 Warning
I need this No GPU darknet.exe because sometime I need to use in the CPU only computer.

[AlexeyAB]

@leekwunfung817

1. GPU (darknet.sln)
   Finally Build Success

Good.

2. No GPU (darknet_no_gpu.sln)
   156 Errors 1400 Warning

Try to build as Release

Try to use OpenCV 3.3.0: https://sourceforge.net/projects/opencvlibrary/files/opencv-win/3.3.0/opencv-3.3.0-vc14.exe/download

There was C API bug in the latest OpenCV-versions.

[leekwunfung817]

Without GPU

• Windows 10

Install Step (by Order):

1. install Visual Studio 2015 Update 3

• https://download.microsoft.com/download/4/8/f/48f0645f-51b6-4733-b808-63e640cddaec/vs2015.3.exe

2. download OpenCV 3.3.0

• https://nchc.dl.sourceforge.net/project/opencvlibrary/opencv-win/3.3.0/opencv-3.3.0-vc14.exe

Complie Step:

1. Set project

• Release
• x64

(Right click darknet_no_gpu project) > (Click Properties)
2. Configuration Properties > C/C++ > Additional include Directories
add

• D:\opencv_3.3.0\build\include
• D:\opencv_3.3.0\build\include\opencv
• D:\opencv_3.3.0\build\include\opencv2

3. Configuration Properties > Linker > Input > Additional Dependencies
   add

• D:\opencv_3.3.0\build\x64\vc14\lib\opencv_world330.lib
• D:\opencv_3.3.0\build\x64\vc14\lib\opencv_world330d.lib

Please use in the MS-DOS command

darknet-master\build\darknet\x64\darknet_no_gpu.exe

Build Success

[AlexeyAB]

@leekwunfung817

Without GPU
Build Success

Excellent!

[leekwunfung817]

I may share the Windows 10 installation step to another people, avoid the other to waste time.

With NVIDIA GPU

• Windows 10
• GPU NVIDIA GeForce GT 710

Install Step (by Order):

1. install Visual Studio 2015 Update 3

• https://download.microsoft.com/download/4/8/f/48f0645f-51b6-4733-b808-63e640cddaec/vs2015.3.exe

2. download OpenCV 3.4.0

• https://nchc.dl.sourceforge.net/project/opencvlibrary/opencv-win/3.4.0/opencv-3.4.0-vc14_vc15.exe

3. install NVIDIA Cuda 9.1

• https://developer.download.nvidia.com/compute/cuda/9.1/secure/Prod/local_installers/cuda_9.1.85_win10.exe?nCAjZE0jUxSE-AcJXMlPk4A3evyLV-gHoreH1tYsooan2BDHNIbM2c4hwYCsVz1DxuPIrVqCqffEipmpGI9EBoXGzuX5C9A05yvvN6S0HBWexrtupkQlXlDGdSGu-5znSfriBqbbPzDWdMBiWc8n-9jl2rC7_6oCI2BYGDX1a1fL-Q

4. download NVIDIA CuDNN 7

• https://developer.download.nvidia.com/compute/machine-learning/cudnn/secure/v7.0.5/prod/9.1_20171129/cudnn-9.1-windows10-x64-v7.zip?gLSaZu2MVANj6BmRhbWbgraHeB4D67aB_223v2QqgdM60BxkRCpWK5wmi2TE8-3FjaTdIxJwabY_wakOSogwnRZIcS5dcnG3T0SZ3GnbKR7Rc_VlLmaU792mgc7d4t6ro0_4OZiVqeysiKOeNLv8_LjC9B7-kwl8sISwXgVpp23qr80UdZ-wpDPhU7txxQhW2sh1A8FFJhdxbYX9

Complie Step:
put CuDNN to NVIDIA GPU Computing Toolkit
setup the darknet.sln properties

• (cudnn-9.1-windows10-x64-v7\cuda\bin\cudnn64_7.dll) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\bin)

• (cudnn-9.1-windows10-x64-v7\cuda\lib\x64\cudnn.lib) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\lib\x64)

• (cudnn-9.1-windows10-x64-v7\cuda\include\cudnn.h) to (C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.1\include)

1. Set project

• Release
• x64
  

(Right click darknet project) > (Click Properties)

2. Configuration Properties > C/C++ > Additional include Directories
   add

• D:\opencv_3.4.0\build\include
• D:\opencv_3.4.0\build\include\opencv
• D:\opencv_3.4.0\build\include\opencv2
  

3. Configuration Properties > Linker > Input > Additional Dependencies
   add

• D:\opencv_3.4.0\build\x64\vc14\lib\opencv_world340.lib
• D:\opencv_3.4.0\build\x64\vc14\lib\opencv_world340d.lib
  

3. Configuration Properties > CUDA C/C++ > CUDA Toolkit Custom Dir
   add

• C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA
  

Please use in the MS-DOS command
darknet-master\build\darknet\x64\darknet.exe

1. GPU (darknet.sln)
Finally Build Success
But two warnings

Without GPU

• Windows 10

Install Step (by Order):

1. install Visual Studio 2015 Update 3

• https://download.microsoft.com/download/4/8/f/48f0645f-51b6-4733-b808-63e640cddaec/vs2015.3.exe

2. download OpenCV 3.3.0

• https://nchc.dl.sourceforge.net/project/opencvlibrary/opencv-win/3.3.0/opencv-3.3.0-vc14.exe

Complie Step:

1. Set project

• Release
• x64
  

(Right click darknet_no_gpu project) > (Click Properties)
2. Configuration Properties > C/C++ > Additional include Directories
add

• D:\opencv_3.3.0\build\include
• D:\opencv_3.3.0\build\include\opencv
• D:\opencv_3.3.0\build\include\opencv2
  

3. Configuration Properties > Linker > Input > Additional Dependencies
   add

• D:\opencv_3.3.0\build\x64\vc14\lib\opencv_world330.lib
• D:\opencv_3.3.0\build\x64\vc14\lib\opencv_world330d.lib
  

Please use in the MS-DOS command
darknet-master\build\darknet\x64\darknet_no_gpu.exe

Build Success