pytorch / pytorch/executorch

ExecuTorch Test Infrastructure

Open
#14,023 0 comments 0 reactions 1 assignee View on GitHub

@GregoryComer is already working on this.

Since Sep 5, 2025.

backend tester
Dominant language
Python
Stars
5k
Forks
1.2k
Avg merge
2d 10h
Merged PRs (30d)
581

Description

Summary

This is the top-level tracking task for work regarding the ExecuTorch Test Infrastructure (Backend Tester, name TBD).

Tasks

  • Split model test jobs into one job per model.
  • Expose error messages in markdown summary.
  • Clean up remaining operators tests failing for reasons outside of backend control or reasonable expectations.
  • Move each backend test job into a dedicated workflow. This will make it easier to find specific results.
  • Run a small smoke test on the test infrastructure on pull. This will prevent breakages.
  • Look into displaying latest summary results on the HUD or other location.
  • Investigate using pytest-xdist to recover from native crashes. They currently end the entire test run due to the unrecoverable nature of native crashes / memory corruption.
  • Create a summary job to accumulate results for each backend and display top-level metrics (pass percentage for f32 models, operator pass percent).

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.