Skip to content

Latest commit

 

History

History
223 lines (166 loc) · 14.1 KB

README.md

File metadata and controls

223 lines (166 loc) · 14.1 KB
Raw

Join the chat at https://gitter.im/Microsoft/CNTK

Latest news

2018-01-22. CNTK support for CUDA 9

CNTK now supports CUDA 9/cuDNN 7. This requires an update to build environment to Ubuntu 16/GCC 5 for Linux, and Visual Studio 2017/VCTools 14.11 for Windows. With CUDA 9, CNTK also added a preview for 16-bit floating point (a.k.a FP16) computation.

Please check out the example of FP16 in ResNet50 here

Notes on FP16 preview:

  • FP16 implementation on CPU is not optimized, and it's not supposed to be used in CPU inference directly. User needs to convert the model to 32-bit floating point before running on CPU.
  • Loss/Criterion for FP16 training needs to be 32bit for accumulation without overflow, using cast function. Please check the example above.
  • Readers do not have FP16 output unless using numpy to feed data, cast from FP32 to FP16 is needed. Please check the example above.
  • FP16 gradient aggregation is currently only implemented on GPU using NCCL2. Distributed training with FP16 with MPI is not supported.
  • FP16 math is a subset of current FP32 implementation. Some model may get Feature Not Implemented exception using FP16.
  • FP16 is currently not supported in BrainScript. Please use Python for FP16.

To setup build and runtime environment on Windows:

  • Install Visual Studio 2017 with following workloads and components. From command line (use Community version installer as example): vs_community.exe --add Microsoft.VisualStudio.Workload.NativeDesktop --add Microsoft.VisualStudio.Workload.ManagedDesktop --add Microsoft.VisualStudio.Workload.Universal --add Microsoft.Component.PythonTools --add Microsoft.VisualStudio.Component.VC.Tools.14.11
  • Install NVidia CUDA 9
  • From PowerShell, run: DevInstall.ps1
  • Start VCTools 14.11 command line, run: cmd /k "%VS2017INSTALLDIR%\VC\Auxiliary\Build\vcvarsall.bat" x64 --vcvars_ver=14.11
  • Open CNTK.sln from the VCTools 14.11 command line. Note that starting CNTK.sln other than VCTools 14.11 command line, would causes CUDA 9 build error.

To setup build and runtime environment on Linux using docker, please build Unbuntu 16.04 docker image using Dockerfiles here. For other Linux systems, please refer to the Dockerfiles to setup dependent libraries for CNTK.

2017-12-05. CNTK 2.3.1 Release of Cognitive Toolkit v.2.3.1.

CNTK support for ONNX format is now out of preview mode.

If you want to try ONNX, you can build from master or pip install one of the below wheel that matches you Python environment.

For Windows CPU-Only:

For Windows GPU:

For Windows GPU-1bit-SGD:

Linux CPU-Only:

Linux GPU:

Linux GPU-1bit-SGD:

You can also try one of the below NuGet package.

2017-11-22. CNTK 2.3 Release of Cognitive Toolkit v.2.3.

Highlights:

  • Better ONNX support.
  • Switched to NCCL2 for better performance in distributed training.
  • Improved C# API.
  • OpenCV is not required to install CNTK, it is only required for Tensorboard Image feature and image reader.
  • Various performance improvement.
  • Added Network Optimization API.
  • Faster Adadelta for sparse.

See more in the Release Notes.
Get the Release from the CNTK Releases page.

2017-11-10. Switch from CNTKCustomMKL to Intel MKLML. MKLML is released with Intel MKL-DNN as a trimmed version of Intel MKL for MKL-DNN. To set it up:

On Linux:

sudo mkdir /usr/local/mklml
sudo wget https://github.com/01org/mkl-dnn/releases/download/v0.11/mklml_lnx_2018.0.1.20171007.tgz
sudo tar -xzf mklml_lnx_2018.0.1.20171007.tgz -C /usr/local/mklml

On Windows:

Create a directory on your machine to hold MKLML, e.g. mkdir c:\local\mklml
Download the file [mklml_win_2018.0.1.20171007.zip](https://github.com/01org/mkl-dnn/releases/download/v0.11/mklml_win_2018.0.1.20171007.zip).
Unzip it into your MKLML path, creating a versioned sub directory within.
Set the environment variable `MKLML_PATH` to the versioned sub directory, e.g. setx MKLML_PATH c:\local\mklml\mklml_win_2018.0.1.20171007

2017-10-10. Preview: CNTK ONNX Format Support Update CNTK to support load and save ONNX format from https://github.com/onnx/onnx, please try it and provide feedback. We only support ONNX OPs. This is a preview, and we expect a breaking change in the future.

  • Support loading a model saved in ONNX format.
  • Support saving a model in ONNX format, not all CNTK models are currently supported. Only a subset of CNTK models are supported and no RNN. We will add more in the future.

To load an ONNX model, simply specify the format parameter for the load function.

import cntk as C

C.Function.load(<path of your ONNX model>, format=C.ModelFormat.ONNX)

To save a CNTK graph as ONNX model, simply specify the format in the save function.

import cntk as C

x = C.input_variable(<input shape>)
z = create_model(x)
z.save(<path of where to save your ONNX model>, format=C.ModelFormat.ONNX)

If you want to try ONNX, you can build from master or pip install one of the below wheel that matches you Python environment.

For Windows CPU-Only:

For Windows GPU:

Linux CPU-Only:

Linux GPU:

2017-09-25. CNTK September interation plan posted here.

2017-09-24. CNTK R-binding now available here.

2017-09-15. CNTK 2.2
Release of Cognitive Toolkit v2.2.

Hightlights:

  • NCCL 2 support
  • New learner interface
  • A C#/.NET API that enables people to build and train networks
  • New C++ and C# eval examples
  • New nodes
  • Tensorboard image support for CNTK

See more in the Release Notes.   Get the Release from the CNTK Releases page.

See all news

Introduction

The Microsoft Cognitive Toolkit (https://cntk.ai), is a unified deep-learning toolkit that describes neural networks as a series of computational steps via a directed graph. In this directed graph, leaf nodes represent input values or network parameters, while other nodes represent matrix operations upon their inputs. CNTK allows to easily realize and combine popular model types such as feed-forward DNNs, convolutional nets (CNNs), and recurrent networks (RNNs/LSTMs). It implements stochastic gradient descent (SGD, error backpropagation) learning with automatic differentiation and parallelization across multiple GPUs and servers. CNTK has been available under an open-source license since April 2015. It is our hope that the community will take advantage of CNTK to share ideas more quickly through the exchange of open source working code.

Installation

Learning CNTK

You may learn more about CNTK with the following resources:

More information

Disclaimer

CNTK is in active use at Microsoft and constantly evolving. There will be bugs.

Microsoft Open Source Code of Conduct

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.