mlcommons / mlcommons/inference

Inference 5.0 submission failure for Llama3.1-405b caused by hardcode in submission checker

Open
#2,142 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.6k
Forks
650
Avg merge
1d 22h
Merged PRs (30d)
6

Description

In the submission checker, this is a hardocded value here:
inference/tools/submission/submission_checker.py at master · mlcommons/inference

and this is in the official user.conf in llama3, which is lower than that number
inference/language/llama3.1-405b/user.conf at master · mlcommons/inference
and the packaging will fail

Please look into this.

Thanks,
Zhong
AMD

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

Start in tools/submission/submission_checker.py around line 598 and compare the hardcoded value with language/llama3.1-405b/user.conf around line 9. Reproduce the submission packaging failure, then verify that the checker accepts the official llama3.1-405b configuration without weakening the intended validation.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
2/5
Estimated time
1-3 hours
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
45/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.