Skip to content

Latest commit

 

History

History
75 lines (56 loc) · 5.68 KB

release-notes.md

File metadata and controls

75 lines (56 loc) · 5.68 KB
Raw

Release Notes

0.5.0.755748 (2018-09-02)

  • Fix an issue that prevents users from importing certain ONNX model.

0.5.0.752675 (2018-08-30)

  • Fix an issue that the most recent version of Visual Studio cannot install the extension.

0.5.0.747136 (2018-08-25)

  • When first starting Tools for AI, an installation page is shown for guiding local AI development environment setup. Users can launch a one-click installer from this page to install required software packages into the default Python environment.
  • We design a new start page to help users build first AI application / train first AI Model within 3 steps. The start page also contains useful AI examples and development resources.
  • Azure Cognitive Services is organized by subscription in the Server Explorer. When creating a cognitive service, users can specify existing resource group or create a new one.
  • Docker image list is updated for job submission to remote machines. And a new all-in-one docker image including all popular AI / DL frameworks is provided.
  • Besides sending feedback in GitHub, now user can send feedback via Email too. Moreover, a more proactive feedback channel is built - will prompt user to give feedback when some conditions met.
  • Azure Batch AI support is temporarily removed, will get the support of Batch AI back soon.
  • Telemetry and stability improvements.

0.4.1.664470 (2018-06-11)

Please install recent version of Visual Studio and upgrade CUDA / AI frameworks.

Since setting up deep learning and machine learning software as well as their dependencies is not an easy task, we recommend that you use the one-click installer to install them automatically.

  • Initial Simplified Chinese language support.
  • CUDA 9.0 and recent version of AI frameworks support.
  • Bug fixes and stability improvements.

0.4.0.620119 (2018-05-06)

We add many cool features in this release, which requires recent version of Visual Studio and upgrading CUDA / AI frameworks.

Since setting up deep learning and machine learning software as well as their dependencies is not an easy task, we recommend that you use the one-click installer to install them automatically.

  • CUDA 9.0 and recent version of AI frameworks support.
  • Infuse AI into your apps with Microsoft Cognitive Services to see, hear, speak, understand and interpret your user needs through natural methods of communication.
    • Add Cognitive Services via GUI wizard and manage them in Visual Studio.
    • Build Cognitive Service apps from pre-defined project templates.
    • Train and manage your own Custom Vision model from the generated training project or web portal.
  • Build intelligent apps using pre-trained AI models with the Microsoft.ML.Scoring library
    • Once training is complete, building intelligent apps in Visual Studio is as easy as putting your trained model in your app just like any other library or resource.
    • Visual Studio Tools for AI makes this easy by enabling you to create a Model Inference Library project which automatically optimizes your ONNX/TensorFlow model for serving, as well as placing on optimized ONNX/TensorFlow runtime in the project.
    • Moreover, Visual Studio Tools for AI generates a C# stub class to simplify interaction with models in your app.
    • These Model Inference Library projects can be further deployed as NuGet packages for convenient distribution.
  • Interoperation between different AI frameworks through model file conversion.
    • Convert Core ML, TensorFlow, scikit-learn, XGBoost and LIBSVM models to ONNX format.
    • Integration with WinMLTools, ONNXMLTools and tf2onnx converters.
  • View network architecture and parameters of AI models in your training and inference projects. Please install the latest version of Netron.
  • Run remote machine jobs in Docker.
  • Log in automatically to your password-based remote machines (e.g. Azure DSVM/DLVM). Please install the latest version of Putty, and set the directory of putty.exe to %Path% environment variable.
  • Add PyTorch application template.
  • Experimental support for Open Platform for AI (PAI).
    • Submit training projects to PAI clusters.
    • Manage jobs and files with GUI tools.
  • Bug fixes and stability improvements.

0.3.3.0 (2018-02-10)

  • Improve the experience of creating Azure ML projects from existing Visual Studio projects.
  • Improve the UI layout and parameter validation of Azure Batch AI cluster creation and job submission to keep consistent with Azure web portal.
  • Connect to a remote machine with OpenSSH client.
  • Bug fixes and stability improvements.

0.3.2.0 (2017-12-19)

  • Check extension dependencies such as Python Tools, and automatically install them if necessary for VS 2017 extension.
  • Refine the job submission experiences of Azure Batch AI, and support NFS servers.
  • Bug fixes and stability improvements.

0.3.0.0 (2017-11-17)

  • Initial version.