Clarke Rahrig

A tall, Texan, techie working @Microsoft on #WindowsAI and #DirectML, @CockrellSchool @UTAustin alum 🤘

[DirectX 12 @DirectX12]

Tomorrow during #MSBuild, we’re featuring a demo of Stable Diffusion running locally with hardware acceleration that's both blazingly fast and cross-platform! https://t.co/PTqvUlbe4j

[DirectX 12 @DirectX12]

We’ve got some exciting DirectML demos and announcements at #MSBuild this year! Digital passes are completely free, so make sure you don’t miss our session: https://t.co/PTqvUlbe4j

[DirectX 12 @DirectX12]

DirectML is here for TensorFlow 2.9!!

Offering DirectML as a PluggableDevice for TensorFlow means developers can accelerate their TF training workloads on any DX12 GPU. #TensorFlow #DirectX

Check out the blog for more info - https://t.co/5p8WWrXNTB

[DirectX 12 @DirectX12]

Today we released another preview of PyTorch-DirectML with increased support for training computer vision models. Learn more here: https://t.co/tC4kuhIBhR #DirectX #PyTorch

[Craig Loewen @craigaloewen]

WSL is coming pre-installed on select upcoming HP workstations for data science and machine learning!

"WSL 2 comes pre-installed and pre-enabled on select Z by HP workstations, so users can quickly and easily maximize productivity out of the box."

https://t.co/KyBleG96Jh

[Clarke Rahrig @crahrig]

The initial preview of PyTorch-DirectML has arrived! 🎉

Try it out by running "pip install pytorch-directml" in your Python environment of choice. Read on about the #DirectML backend for #PyTorch brining GPU accelerated ML training to any @DirectX12 GPU:

https://t.co/YYeQeSLXPs

[Windows Developer @windowsdev]

The Windows AI team is excited to announce the first preview of DirectML as a backend to PyTorch for training ML models! Head over to our blog to learn more: https://t.co/xyFl21s8cX

[Windows Developer @windowsdev]

Check out how #Windows11 has made support for GPU accelerated machine learning training within the Windows Subsystem for Linux broadly available: https://t.co/rfQPiTy8eX

[Clarke Rahrig @crahrig]

GPU accelerated ML workflows in #WSL are now broadly available with the release of #Windows11! 🥳

Check out the link for details on setting up your existing ML workflows with #CUDA or #DirectML

https://t.co/pnQJ0kOFgK

[Windows Developer @windowsdev]

Over the past year, we launched the TensorFlow-DirectML preview for Windows and WSL, working with the TensorFlow community. Today, we’re excited to announce our first generally consumable package of TensorFlow-DirectML 🥳

Read more here: https://t.co/RLauOrfUUT

[Windows Developer @windowsdev]

Are you learning about TensorFlow using Jupyter Notebooks in #WSL? Here's how you can get GPU acceleration using #TensorFlow with #DirectML.
https://t.co/QW3DCYRJhb

[Clarke Rahrig @crahrig]

Super excited to announce the Windows AI Platform Blog – a developer blog focused on providing in-depth views at new AI features and materials to help you start building AI experiences today!

https://t.co/YNHOqJmMHR

[Clarke Rahrig @crahrig]

@gpu3d @gpu3d GUI app support is now in the latest Windows Insiders builds! Since you're already setup with CUDA in WSL it'll only take a couple other updates to enable this support.

You can find all the details here (cc: @craigaloewen): https://t.co/WNamimrVN8

[Clarke Rahrig @crahrig]

@uguralikaplan There is not a date yet but stay tuned for more details to come on this front!

[Clarke Rahrig @crahrig]

GPU support for ML training in #WSL is here! 🎉🎊🥳

Check out https://t.co/9BhIFr8bUk for details on how to get started with your ML workflows TODAY on #CUDA or #DirectML! We're looking forward to your feedback on #gpuinwsl

[Windows Developer @windowsdev]

Drum roll, please 🥁🥁🥁

We are excited to announce the public release of DirectML as a standalone API for Win32, UWP, and WSL applications in a single NuGet package ▶ Microsoft .AI.DirectML ◀️

https://t.co/w126l1gWqN

[Clarke Rahrig @crahrig]

Check out the latest from #DirectML and see how to bring your AI workload to any GPU!

https://t.co/rKNSfC6KE9

[Clarke Rahrig @crahrig]

@arjunkrishna Hi Arjun! Stay tuned for more details on this front as we're still focused on building on feedback from the preview. What have you found the most useful?

[Clarke Rahrig @crahrig]

@craigaloewen @sferquel

[Lobe @lobe_ai]

Today we are excited to launch a public preview of the new Lobe! Everything you need to train custom machine learning models in a free, easy to use app. Check it out! https://t.co/JSYrSacO88

[Clarke Rahrig @crahrig]

@madrid_ankit If you're setting up tensorflow-directml all you need to do is install in your Python env of choice, which could be miniconda. DirectML is in the tf-dml package and as you said, provides the GPU acceleration.

Today tf-dml only supports TF1.15. Do you use a different ver of TF?

[Clarke Rahrig @crahrig]

Check out the latest episode of Tabs vs Spaces to learn about setting up the GPU acceleration in WSL preview.

I walkthrough configuring NVIDIA #CUDA and the #TensorFlow with #DirectML package to use with your ML workflows!

https://t.co/VZMWhBcOEv