fix: recover PR 1888 ingest via sweep reuse

[Oseltamivir]

Recover the missing official ingest for PR #1888 (merge e72ade4e).

The push-to-main Run Sweep run 28027569383 did not reuse the validated PR sweep because /reuse-sweep-run was not authorized before merge. Its GPU benchmark jobs were cancelled and trigger-ingest fired without valid benchmark data, so the official ingest of PR #1888's results did not complete.

This recovery reuses the validated PR sweep run 28009839213 (source head d567f256, attempt 1) to perform the official ingest.

---

Note

Low Risk
Documentation-only changelog append for ingest recovery; no runtime, auth, or benchmark configuration changes.

Overview
Recovers the official benchmark ingest for GB300 MiniMax-M3 FP8 Dynamo-vLLM (minimaxm3-fp8-gb300-dynamo-vllm) that failed when PR #1888 merged without an authorized /reuse-sweep-run, so post-merge sweep jobs were cancelled and ingest had no valid GPU results.

The change appends a new perf-changelog.yaml entry (same description as #1888: ARM64 perf image minimax-m3-perf-arm64-13.0.1-7a67223, FlashInfer MNNVL workspace) with pr-link set to this recovery PR (#1897). Benchmark recipes and settings are unchanged; validated PR sweep run 28009839213 is reused instead of re-running GPUs.

^{Reviewed by Cursor Bugbot for commit a9fdb3b. Bugbot is set up for automated code reviews on this repo. Configure here.}

[Oseltamivir]

/reuse-sweep-run 28009839213

[github-actions]

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

• https://docs.sglang.io/cookbook/intro
• https://recipes.vllm.ai/

If it is not, please create a PR first before we can merge your single node 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

• https://github.com/vllm-project/recipes
• https://github.com/sgl-project/sglang/tree/main/docs_new

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.

---

感谢你的贡献！对于 vLLM 与 SGLang，请确保你的 recipe 与官方 vLLM recipes 和/或 SGLang cookbook 保持一致

• https://docs.sglang.io/cookbook/intro
• https://recipes.vllm.ai/

如果不一致，请先创建一个 PR，之后我们才能将你的单节点 PR 合并到 master 分支。让我们确保文档保持一流水准，使整个 ML 社区都能从你的辛勤工作中受益！谢谢

• https://github.com/vllm-project/recipes
• https://github.com/sgl-project/sglang/tree/main/docs_new

PR 作者有责任确保合并后所有 GitHub Action 任务完全通过。 很多时候失败只是偶发抖动（flake），重新运行失败的任务即可解决。如果选择重新运行失败的任务，PR 作者有责任确保其最终通过。参见 GitHub 关于重新运行失败任务的文档：https://docs.github.com/en/actions/how-tos/manage-workflow-runs/re-run-workflows-and-jobs#re-running-failed-jobs-in-a-workflow

一般而言，PR 作者应先向相应公司的 CODEOWNERS 请求审阅并获得 PR 批准，然后再请求核心维护者审阅。

如需更多帮助，PR 作者可通过 Slack 联系核心维护者。

[Oseltamivir]

✅ Ingest recovery complete

Official ingest for PR #1888 (minimaxm3-fp8-gb300-dynamo-vllm) has been recovered via sweep reuse.

Root cause

The push-to-main Run Sweep run 28027569383 on merge e72ade4e did not reuse the validated PR sweep because /reuse-sweep-run was not authorized before merge. Its reuse-ingest-artifacts job was skipped, the GPU benchmark jobs ran and were cancelled, and trigger-ingest then fired (run conclusion cancelled) without valid benchmark data — so the official ingest of PR #1888's results never landed.

Recovery chain

Stage	Reference
Failed target (push, no reuse)	run 28027569383 on merge e72ade4e — cancelled
Reused source (PR sweep)	run 28009839213, attempt 1, head d567f256 — success
Recovery PR	#1897 → squash merge ca792660
Recovery push run (reuse)	run 28028703404 — success, reuse-enabled=true
Downstream DB ingest	InferenceX-app run 28028786577 — success

Recovered scope & artifacts

• Config key: minimaxm3-fp8-gb300-dynamo-vllm (MiniMax-M3 MXFP8, GB300, Dynamo-vLLM)
• Reused result set: 28 fixed-sequence benchmark rows + 8 eval rows (0 agentic), covering multi-node 1k1k/8k1k and multinode evals
• Source artifacts passed validate_reusable_sweep_artifacts.py (no duplicate rows, run stats present, raw/aggregate eval consistent)

Verification

• Recovery push run: setup ran, reuse-ingest-artifacts validated, upload-changelog-metadata + trigger-ingest succeeded; all GPU/collect jobs skipped.
• Reuse resolved to source run 28009839213 attempt 1 (Run ID: 28009839213 .../attempts/1), triggered by recovery run 28028703404.
• Ingest stages all succeeded: artifact download → flatten → DB ingest (ingest-ci-run complete) → apply run overrides → database verification → Vercel cache invalidation → unmapped-entity check (no unmapped entities; Slack failure/unmapped notifications skipped).