Microsoft/CNTK

Microsoft Cognitive Toolkit (CNTK), an open source deep-learning toolkit http://cntk.ai

Mukundh Bhushan

@MukundhBhushan

This is the git hub repo: https://github.com/NebutechOpenSource/MLGen

Thank you

Michael Coen

@mhcoen

Hi all. We've been using CNTK with Cuda Toolkit v10.0 and cuDNN v7.5.1. I just tried CNTK with the latest releases of the CUDA Toolkit and cuDNN library for Cuda 10.2. CNTK works fine with these but I was expecting I might see a performance improvement during training given some of NVidia's claims. However, benchmark scores during training are unchanged. Any thoughts here? Was my expectation unreasonable? Just wondering what I'm not getting. (The testing was done on a GTX 1080. I have a newer (but smaller) Quadro T2000 in my laptop that I haven't upgraded for CUDA 10.2 yet. Any reason to expect I'd see an improvement with a different GPU?)

Many thanks.

ambarchakrabarty-DnA

@ambarchakrabarty-DnA

Hi I am looking for an AI solution that can automatically read data from excel, add conditional logic to excel data currently being added to manually to excel and then finally load data into MS Access . Can I use CNTK to deliver this solution ?

delzac

@delzac

@ambarchakrabarty-DnA No cntk is a deep learning framework and doesn't deal with excel. You might want to try PyXLL

ambarchakrabarty-DnA

@ambarchakrabarty-DnA

@delzac many thanks for getting back . Sure, I will have a look. Thanks !

JamesLear92

@JamesLear92

Hi, can someone help me understand why CNTK is not getting further releases? There is nothing else that can compete when it comes to C# development.
What should C# developers use instead? Tensorflow.Net is not as C# friendly, and won't work at all with a GTX 760

Haiping

@Oceania2018

@JamesLear92 There is a link how to use GPU for ML.NET and TF.NET: https://devblogs.microsoft.com/dotnet/announcing-ml-net-1-4-global-availability-machine-learning-for-net/

Syed Farhan Ahmad

@born-2learn

Will CNTK be participating as an org in Google Summer of Code 2020 by any chance?

Phuong Ngoc Chau

@ph0123

Can I use rpc, MPI_send, MPI_recev with CNTK? where I can find basic example distributed CPU with CNTK?

Thanks

Dhiraj K

@DhirajK32398669_twitter

Hi All

I am tying to install cntk on my windows 10 machine

by pip install cntk

but it is giving error

ERROR: Could not find a version that satisfies the requirement cntk (from versions: none)
ERROR: No matching distribution found for cntk

how to fix this

delzac

@delzac

@DhirajK32398669_twitter What version of python are you using?

Install instructions from here

https://docs.microsoft.com/en-us/cognitive-toolkit/setup-windows-python?tabs=cntkpy26

maddhatters-dev

@maddhatters-dev

@DhirajK32398669_twitter , I ran into that problem for a while, you can download it manually, more often then not you will need to run it after downloading in your sdk, many python packages are downloads that get unpacked, not auto installed for security reasons. Make sure you download the one for your operating system https://github.com/microsoft?q=cog&type=&language=

newoneincntk

@newoneincntk

Hello, any guys here?
I use PIP to install CNTK on WIN10 1909/1903. But when I run a script, although there is no error message, the training process does not start.
My PC is AMD 3950x + 2080TI.

Qian Xin

@Neteraxe

Is it the support of python 3.8.3 available?

ERROR: Could not find a version that satisfies the requirement cntk (from versions: none)
ERROR: No matching distribution found for cntk

delzac

@delzac

@Neteraxe No ther's no support for python 3.8 only up to python 3.6

Иван Сердюк

@oceanfish81_twitter

Is there any interest to support Clang, for Windows builds?

I was thinking about clang-cl , for the C++ code (P/Invoke related) - and OpenCL, GPGPU code could be compiled by existing Clang (11.0 RC2)

Иван Сердюк

@oceanfish81_twitter

I have some issues here:
\Source\Math>msbuild Math.vcxproj /p:Configuration=Release /p:Platform=x64
Microsoft (R) Build Engine version 16.7.0+b89cb5fde for .NET Framework
Copyright (C) Microsoft Corporation. All rights reserved.

Build started 07.09.2020 12:50:34.
Project "C:\Users\clang\cntk_msvc\Source\Math\Math.vcxproj" on node 1 (default targets).
InitializeBuildStatus:
Touching "C:\Users\clang\cntk_msvc\x64.build\Release\Source\Math\Math.tlog\unsuccessfulbuild".
ClCompile:
All outputs are up-to-date.
C:\Program Files\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.16.27023\bin\HostX86\x64\CL.exe /c /I"C:\Pro
gram Files\Intel\mklml\include" /IC:\Users\clang\cntk_msvcSource\Common\include /I\include /Zi /nologo /W4 /WX /diagn
ostics:classic /sdl /MP /O2 /Oi /Ot /D NDEBUG /D USE_MKL /D NO_SYNC /D WIN32 /D _WINDOWS /D _USRDLL /D MATH_EXPORTS /
D _WINDLL /D _UNICODE /D UNICODE /D NDEBUG /D "CNTK_VERSION=\"2.7\"" /D "CNTK_VERSION_BANNER=\"2.7+\"" /D "CNTK_COMPO
NENT_VERSION=\"2.7\"" /D HAS_MPI=1 /Gm- /EHsc /MD /GS /Gy /Qpar /fp:fast /fp:except- /Zc:wchar_t /Zc:forScope /Zc:inl
ine /openmp /Yu"stdafx.h" /Fp"C:\Users\clang\cntk_msvc\x64.build\Release\Source\Math\Cntk.Math-2.7.pch" /Fo"C:\Users
\clang\cntk_msvc\x64.build\Release\Source\Math\" /Fd"C:\Users\clang\cntk_msvc\x64.build\Release\Source\Math\vc141.
pdb" /Gd /TP /wd4819 /FC /errorReport:queue /d2Zi+ /bigobj BatchNormalizationEngine.cpp ConvolutionEngine.cpp CPUMatr
ixDouble.cpp CPUMatrixFloat.cpp CPUMatrixHalf.cpp CPUMatrixTensorDouble.cpp CPUMatrixTensorFloat.cpp CPUMatrixTensorH
alf.cpp MatrixQuantizerCPU.cpp MatrixQuantizerImpl.cpp Matrix.cpp QuantizedMatrix.cpp
BatchNormalizationEngine.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
ConvolutionEngine.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
CPUMatrixDouble.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
Unknown compiler version - please run the configure tests and report the results
CPUMatrixFloat.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
Unknown compiler version - please run the configure tests and report the results
CPUMatrixHalf.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
Unknown compiler version - please run the configure tests and report the results
CPUMatrixTensorDouble.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
CPUMatrixTensorFloat.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note: see declaration of 'half'
CPUMatrixTensorHalf.cpp
c:\users\clang\cntk_msvc\source\math\half.hpp(36): error C2371: 'half': redefinition; different basic types [C:\Users\c
lang\cntk_msvc\Source\Math\Math.vcxproj]
c:\program files\cuda-11\include\cuda_fp16.hpp(2270): note

Richerd Namsseng

@namsseng

Hello. I have a question. I want using CUDA 11 for CNTK. Becouse I am using RTX 3090.

Dae Hyun Baik

@dh.baik_gitlab

You need to rebuild CNTK with CUDA 11. but it's not piece of cake.

Иван Сердюк

@oceanfish81_twitter

I think Microsoft will always be behind CUDA releases - and target the previous version of it, on any specific date

Where communities thrive

People

Repo info

Activity