Deep Learning Reference Stack

The Deep Learning Reference Stack is an integrated, highly-performant open source stack optimized for Intel® Xeon® Scalable platforms.

Highly-tuned and built for cloud native environments, the release of DLRS enables developers to quickly prototype by reducing complexity associated with integrating multiple software components, while still giving users the flexibility to customize their solutions.

The stack includes highly tuned software components across the operating system (Ubuntu* or Centos*), deep learning framework (TensorFlow*, PyTorch*), deep learning libraries ([Intel® oneAPI Deep Neural Network Library (oneDNN)](https://01.org/dnnl)) and other software components. This open source community release is part of an effort to ensure AI developers have easy access to all features and functionality of Intel platforms.To offer more flexibility, there are multiple versions of the Deep Learning Reference Stack.

DLRS Release Announcement and Performance Reports

• DLRS V0.9.0 release announcement.

• DLRS V0.8.0 release announcement.

• DLRS V0.7.0 release announcement.

• DLRS V0.6.0 release announcement.

• DLRS V0.5.0 release announcement.

DLRS Guides

• Deep Learning Reference Stack Guide
• State-of-the-art BERT Fine-tune training and Inference

DLRS Releases

DLRS with TensorFlow*

• Deep Learning Reference Stack with Tensorflow and Intel® oneAPI Deep Neural Network Library (oneDNN)

DLRS with TensorFlow Serving*

• Build instructions

DLRS with PyTorch*

• Deep Learning Reference Stack with Pytorch and Intel® oneAPI Deep Neural Network Library (oneDNN)

DLRS ML Compiler

• Stacks Deep Learning Compiler