pytorch / pytorch/executorch

Enable workflow time profiling

Open
#8,244 0 comments 0 reactions 1 assignee View on GitHub

@guangy10 is already working on this.

Since Feb 6, 2025.

module: benchmark triaged
Dominant language
Python
Stars
5k
Forks
1.2k
Avg merge
2d 10h
Merged PRs (30d)
581

Description

🚀 The feature, motivation and pitch

We start seeing some workflows running over 60mins on macOS runners recently, which took from 15-30 mins previously. It's hard to tell if the issue is due to additional setup required on macOS runner (as there is no docker image), e.g. package downloading and installation, etc. or due to some AOT perf regression in our stack.

In this task, I want to add a simple way to profiling the time of executing the benchmark workflow, where we can get more insights where the slowness comes from.

Alternatives

No response

Additional context

No response

RFC (Optional)

No response

cc @huydhn @kirklandsign @shoumikhin

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.