open-mmlab / open-mmlab/mmengine

[Feature] Improve eta calculation when using torch.compile

Open
#1,073 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.5k
Forks
464
PR merge metrics
No merged PRs in 30d

Description

What is the feature?

When using torch.compile to compile the model, the first few iters may cost much more time than normal iters. This will lead to unreliable estimation for eta. So the way to estimate eta time needs to be improved when using torch.compile.

Any other context?

No response

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.

Research direction

No file, test, or entry point is named. Start by locating the ETA calculation and the path used with torch.compile; completion should produce reliable estimates despite slower initial iterations and include coverage for the behavior.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning, performance
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
30/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.