Skip to content

Revert "[WIP] [AMD/ROCM] atom minimaxm2.5 fp4 on mi355x" (#1042)#1239

Merged
cquil11 merged 1 commit intomainfrom
revert-1042-atom-minimaxm2.5-fp4
Apr 30, 2026
Merged

Revert "[WIP] [AMD/ROCM] atom minimaxm2.5 fp4 on mi355x" (#1042)#1239
cquil11 merged 1 commit intomainfrom
revert-1042-atom-minimaxm2.5-fp4

Conversation

@cquil11
Copy link
Copy Markdown
Collaborator

@cquil11 cquil11 commented Apr 30, 2026

Reverts #1042.

Reason: the perf-changelog entries that landed with the original PR were incorrect. Reverting so the branch can be re-merged cleanly.

A new PR will be opened from srok/atom_minimaxm2.5_fp4 after this revert lands.

@github-actions
Copy link
Copy Markdown
Contributor

Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook

If it is not, please create a PR first before we can merge your PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow

As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers.

If additional help is needed, PR authors can reach out to core maintainers over Slack.

1 similar comment
@github-actions
Copy link
Copy Markdown
Contributor

Thanks for the contribution! For vLLM & SGLang, please ensure that your recipes is similar to the official vLLM recipes and/or the SGLang cookbook

If it is not, please create a PR first before we can merge your PR into the master branch. Let's ensure that the documentation is first class such that the entire ML community can benefit from your hard work! Thank you

PR authors are responsible for ensuring that after merging, all GitHub Action jobs fully pass. A lot of the time, failures are just flakes and simply re-running the failed jobs will fix it. If re-running failed jobs is attempted, PR authors are responsible for ensuring it passes. See GitHub's docs on re-running failed jobs: https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow

As a rule of thumb, generally, PR authors should request a review & get a PR approval from the respective companies' CODEOWNERS before requesting a review from core maintainers.

If additional help is needed, PR authors can reach out to core maintainers over Slack.

@cquil11 cquil11 merged commit 28b3165 into main Apr 30, 2026
10 checks passed
@cquil11 cquil11 deleted the revert-1042-atom-minimaxm2.5-fp4 branch April 30, 2026 20:21
Copy link
Copy Markdown
Contributor

@claude claude Bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Clean revert of a benchmark config addition; safe to re-merge from a clean branch as described.

Extended reasoning...

Overview

This PR is a straightforward revert of #1042, removing the minimaxm2.5-fp4-mi355x-atom benchmark config from .github/configs/amd-master.yaml, deleting the corresponding launch script benchmarks/single_node/minimaxm2.5_fp4_mi355x_atom.sh, and dropping the associated perf-changelog.yaml entry. The author explicitly notes the perf-changelog entries from the original PR were incorrect and they intend to re-open from srok/atom_minimaxm2.5_fp4 after this lands.

Security risks

None. The change is confined to AMD MI355X benchmark recipe metadata and a benchmark launch shell script. No auth, crypto, permissions, secrets handling, or runtime/inference-path code is touched.

Level of scrutiny

Low. Reverts of benchmark config / changelog additions are mechanical and easy to verify by inspection — the diff matches the inverse of the original add, and the deleted files are isolated to the AMD MI355X atom benchmarking surface. No CODEOWNER-protected production code is affected.

Other factors

The bug-hunting system found no issues. The only timeline activity is the automated recipe-reminder bot. The motivation (broken changelog entries, re-merge planned) is reasonable and the change is fully self-contained.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

Development

Successfully merging this pull request may close these issues.

1 participant