Issue Station
Sign In
Tools hereToolbox

pytorch - ✅(Solved) Fix Some macOS wheels say macOS 11.0+ but inside there's libomp.dylib with minos 14.0 [2 pull requests, 7 comments, 3 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
pytorch/pytorch#177140•Fetched 2026-04-08 00:21:57
View on GitHub
Comments
7
Participants
3
Timeline
59
Reactions
0
Author
uwu-420uwu-420
Participants
atalmanatalmanmalfetmalfetuwu-420uwu-420
Assignees
yangw-devyangw-dev
Timeline (top)
mentioned ×14subscribed ×14referenced ×8commented ×7

Fix Action

Fixed

PR fix notes

PR #177609: [macOS wheel] Add wheel platform tag vs dylib minos validation

Description (problem / solution / changelog)

Fixes https://github.com/pytorch/pytorch/issues/177140

Summary

Adds a check to validate that the macOS wheel platform tag matches the actual minos (minimum OS version) in the Mach-O binaries.

For example, given a wheel like：

torch-2.12.0.dev20260316-cp313-cp313t-macosx_11_0_arm64.whl

The platform tag claims macosx_11_0_arm64, meaning minos should be 11.0. This script runs otool -l on each installed .dylib and verifies the minos from LC_BUILD_VERSION (or LC_VERSION_MIN_MACOSX) matches 11.0.

Changed files

  • .ci/pytorch/smoke_test/check_wheel_tags.py (added, +236/-0)
  • .ci/pytorch/smoke_test/smoke_test.py (modified, +5/-0)

PR #177903: [macOS wheel] Add wheel platform tag vs dylib minos validation (#177609)

Description (problem / solution / changelog)

Fixes https://github.com/pytorch/pytorch/issues/177140

Summary

Adds a check to validate that the macOS wheel platform tag matches the actual minos (minimum OS version) in the Mach-O binaries.

For example, given a wheel like：

torch-2.12.0.dev20260316-cp313-cp313t-macosx_11_0_arm64.whl

The platform tag claims macosx_11_0_arm64, meaning minos should be 11.0. This script runs otool -l on each installed .dylib and verifies the minos from LC_BUILD_VERSION (or LC_VERSION_MIN_MACOSX) matches 11.0.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/177609 Approved by: https://github.com/atalman

Changed files

  • .ci/pytorch/smoke_test/check_wheel_tags.py (added, +236/-0)
  • .ci/pytorch/smoke_test/smoke_test.py (modified, +5/-0)

Code Example

torch-2.10.0-cp314-cp314t-macosx_14_0_arm64.whl
torch-2.10.0-cp314-cp314-macosx_14_0_arm64.whl
torch-2.10.0-cp313-none-macosx_11_0_arm64.whl
torch-2.10.0-cp313-cp313t-macosx_14_0_arm64.whl
torch-2.10.0-cp312-none-macosx_11_0_arm64.whl
torch-2.10.0-cp311-none-macosx_11_0_arm64.whl
torch-2.10.0-cp310-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp313-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp312-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp311-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp310-none-macosx_11_0_arm64.whl
RAW_BUFFERClick to expand / collapse

🐛 Describe the bug

Basically the title.

I just downloaded torch-2.10.0-cp313-none-macosx_11_0_arm64.whl, extracted the wheel and ran otool -l torch/lib/libomp.dylib and found minos 14.0. Isn't that a problem?

What's going on with the torch wheels for macOS on pypi.org anyway? Is all of this intentional or did something break at some point?

torch-2.10.0-cp314-cp314t-macosx_14_0_arm64.whl
torch-2.10.0-cp314-cp314-macosx_14_0_arm64.whl
torch-2.10.0-cp313-none-macosx_11_0_arm64.whl
torch-2.10.0-cp313-cp313t-macosx_14_0_arm64.whl
torch-2.10.0-cp312-none-macosx_11_0_arm64.whl
torch-2.10.0-cp311-none-macosx_11_0_arm64.whl
torch-2.10.0-cp310-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp313-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp312-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp311-none-macosx_11_0_arm64.whl
torch-2.10.0-2-cp310-none-macosx_11_0_arm64.whl

Versions

torch 2.10.0

The rest should be irrelevant I guess?

cc @seemethere @malfet @atalman @tinglvv @nWEIdia @aditvenk

extent analysis

Fix Plan

Fix: Use the Correct Torch Wheel

To fix the issue, you need to use the correct Torch wheel that matches your macOS version. Based on the provided wheel names, it seems that the correct wheel for macOS 14.0 (M1) is torch-2.10.0-cp314-cp314t-macosx_14_0_arm64.whl.

Steps:

  1. Check your macOS version: Run sw_vers in your terminal to check your macOS version.
  2. Download the correct wheel: Go to the PyPI page for Torch and download the wheel that matches your macOS version. For macOS 14.0 (M1), download torch-2.10.0-cp314-cp314t-macosx_14_0_arm64.whl.
  3. Install the correct wheel: Run pip install torch-2.10.0-cp314-cp314t-macosx_14_0_arm64.whl to install the correct wheel.

Example Code:

# Check macOS version
sw_vers

# Download the correct wheel
curl -O https://files.pythonhosted.org/packages/7d/6b/9f9f3b1e3f6b1c5a4f4f9f9f3b1e3f6b1c5a4f4f9f9f3b1e3f6b1c5a4f4f9f9f3b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3f6b1e3

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#api#ssr#installation#tensor shape#autograd error#prompt template#agent execution#callback error#memory management#API rate limit

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING