PyTorch Release 2.8 | Key Information

Hi Team,

We are kicking off the PyTorch 2.8.0 release cycle and continue to be excited for all the great features from our PyTorch community!

WHEN ARE THE CRITICAL DATES FOR THE PYTORCH 2.8.0 RELEASE?

  • M1: Release Announcement (5/28/25)
  • M2: Land Target PRs in PyTorch repo - DEADLINE (6/20/25)
  • M3.1: Release branch cut (6/23/25)
  • M3.2: Release first RC1 Binary for PyTorch Core (6/24/25)
  • M3.3: Ecosystem libraries cut RC Branch (6/25/25)
  • M4: Release branch finalized, Announce final launch date, new major features published (week of 7/7/25) - Final RC is produced.
  • M4.1: Tutorial drafts submission deadline (7/16/25)
  • M5: External-Facing Content Finalized (7/25/25 )
  • M6: Release Day (7/30/25)

CHANGES TO CLASSIFICATION & TRACKING
As mentioned in this RFC, beginning in release 2.8, there are changes to classification and tracking.

Classification

Beginning with release 2.8, feature submissions will be classified as either Stable (API-Stable) or Unstable (API-Unstable), and the previous classifications of Prototype, Beta and Stable, will no longer be used. The requirements for a feature to be considered stable remain the same, and in the RFC we propose a suggested path to stable.

Tracking

What’s changing?

  • All new features should create an RFC. Steps are: New Issue > Release highlight for Proposed Feature. (see image below)

  • The PyTorch team will only actively track a subset of all features to include in the release blogpost. These features will be tracked from inception (initial RFC) till they are stable, after which they will be untagged and no longer tracked. This will provide continuous visibility / up-to-date status on features that the majority of the community relies on.

What’s not changing?

All features continue to be welcome. If you would like a feature to be included in the release blogpost, please mention it in the “Release highlight for Proposed Feature” issue that you create, and include the release version this is targeted to.
Please also read the blog post for the 2.7 release to get familiar with the marketing format.

If you still have questions after reading the material –or at any time during the process–please don’t hesitate to contact one of us: Gregory Chanan, Alban Desmaison, Andrey Talman,Nikita Shulga, Hamid Shojanazeri, Eli Uriegas and Anita Katahoire.

Cheers,

Team PyTorch
#oneteam

1 Like