Skip to content

PyTorch 2.11 Release Blog

0.5358
Chimera Difficulty Score
a synthesis of Flesch-Kincaid, Coleman-Liau, SMOG, and Dale-Chall readability metrics
We are excited to announce the release of PyTorch® 2.11 (release notes)! The PyTorch 2.11 release features the following changes: - Differentiable Collectives for Distributed Training - FlexAttention now has a FlashAttention-4 backend on Hopper and Blackwell GPUs. - MPS (Apple Silicon) Comprehensive Operator Expansion - RNN/LSTM GPU Export Support - XPU Graph This release is composed of 2723 commi...
The PyTorch 2.11 release presents several noteworthy advancements in distributed deep learning research and advanced training techniques, as well as improvements for Apple Silicon devices, Intel GPUs, and AMD GPUs. However, it is important to consider that with any open-source project, the adoption rate of new features may vary depending on user experiences and compatibility with existing workflows. The Differentiable Collectives feature has the potential to significantly impact distributed deep...