PyTorch release 2.1.1 planning

Dear Team

We will be creating patch PyTorch release 2.1.1.

Following are the tentative dates:

  • 11/3 Cherry-pick post deadline (End of day 5PM PST)
  • 11/6 All RC binaries are ready for testing
  • 11/15 Release 2.1.1 General Availability

Please consider the criteria below and follow the cherry picking process. Only low-risk changes may be cherry-picked from main:

  1. Fixes to regressions against the most recent release (e.g. 2.1.0 for 2.1.1 release; see module: regression issue list)
  2. Low risk critical fixes for: silent correctness, backwards compatibility, crashes, deadlocks, (large) memory leaks
  3. Fixes to new features being introduced in 2.1.0 release
  4. Documentation improvements
  5. Release branch specific changes (e.g. blocking ci fixes, change version identifiers)

Issue tracker for this release can be found here
Issues that are targeted to be included in this release can be found here

Hi team,

Do we expect domain library releases (visoin, audio, text, data, torchserve) to go along with this release as well?
We are packaging PyTorch libraries in AWS Deep Learning Containers (Start via Cloud Partners | PyTorch) and would like to know what will happen for torchserve and data

  1. Will torchserve become a domain library and share the same release cycle as torch like vision and audio?
  2. Torch data is deprecated, do we expect future updates, new tags, new releases?

This helps us plan ahead Thanks!

1 Like

HI @roywei

Yes we should release pytorch, vision, audio, data and text with this patch release. 2.1.1.
For Question #1 let me ask torchserve team to comment. For Question #2 - please stay tuned for release 2.2 announcement. And consult GitHub - pytorch/data: A PyTorch repo for data loading and utilities to be shared by the PyTorch domain libraries. for any changes.

Thank you,

Hi @roywei

Sharing what we discussed in the call.

  • TorchServe release won’t be tightly coupled with PyTorch Release
  • TorchServe version is not tightly coupled with PyTorch version.
  • We will continue to make a TorchServe release after a major PyTorch Release.
  • We will not have a mandatory TorchServe release after every PyTorch minor release. We will do it only if its necessary.

Thank you guys for the clarification and all the help driving the releases!