mirror of
https://github.com/saymrwulf/pytorch.git
synced 2026-05-15 21:00:47 +00:00
fb64caedb5
Summary: https://github.com/pytorch/pytorch/issues/54779 split out the logic from our "Lint" workflow into a separate workflow that allows us to annotate PRs from forks. However, as of https://github.com/pytorch/pytorch/issues/54689, it is possible for the "Lint" workflow to be canceled, in which case it may not upload the "flake8-py3" and "clang-tidy" artifacts that the "Add annotations" workflow expects. This often results in GitHub pointlessly sending notification emails due to the failure in the "Add annotations" workflow. This PR fixes the issue by gracefully handling the case where the expected artifact is absent. Pull Request resolved: https://github.com/pytorch/pytorch/pull/55242 Test Plan: I tested this in the same external sandbox repo used to test https://github.com/pytorch/pytorch/issues/54779. Reviewed By: malfet Differential Revision: D27540120 Pulled By: samestep fbshipit-source-id: 47cc02950edbbc6381033bda2fe4570cb3e331cb |
||
|---|---|---|
| .. | ||
| add_annotations.yml | ||
| auto_label.yml | ||
| build_linux_conda.yml | ||
| build_linux_libtorch.yml | ||
| build_linux_wheels.yml | ||
| cancel_redundant_workflows.yml | ||
| clang_format.yml | ||
| jit_triage.yml | ||
| lint.yml | ||
| push_nightly_docker_ghcr.yml | ||
| quantization_triage.yml | ||
| stale_pull_requests.yml | ||
| test_tools.yml | ||
| update_s3_htmls.yml | ||