Deprecate MiniMax M2.5 / MiniMax M2.7 model architecture benchmark configs

[functionstackx]

Summary

• Remove active MiniMax M2.5 benchmark entries from the AMD and NVIDIA master configs.
• Archive the removed config blocks under .github/configs/deprecated/amd-minimaxm2.5-m2.7-master.yaml and .github/configs/deprecated/nvidia-minimaxm2.5-m2.7-master.yaml.
• Move MiniMax M2.5 benchmark launchers into deprecated folders for fixed-seq-len, agentic, and the MI355X multi-node disagg script.

Note: no MiniMax M2.7 benchmark scripts or active master config keys were present on origin/main; the archive filenames still include M2.7 per the request.

M3 has surpassed M2.5/M2.7 token volume by a lot

Validation

• /tmp/inferencex-validate-py312-m25-m27/bin/python -m pytest utils/matrix_logic/ -v (178 passed)
• /tmp/inferencex-validate-py312-m25-m27/bin/python utils/matrix_logic/generate_sweep_configs.py full-sweep --config-files .github/configs/amd-master.yaml --model-prefix minimaxm2.5 -> 0 entries
• /tmp/inferencex-validate-py312-m25-m27/bin/python utils/matrix_logic/generate_sweep_configs.py full-sweep --config-files .github/configs/nvidia-master.yaml --model-prefix minimaxm2.5 -> 0 entries
• Full-sweep generation completed for active AMD and NVIDIA masters (1213 AMD entries, 2448 NVIDIA entries)

---

Note

Medium Risk
Large-scale change to CI benchmark matrices: it stops all MiniMax M2.5 scheduled runs but does not alter runtime code; mistaken removal of non-M2.5 keys would be the main failure mode.

Overview
Removes all MiniMax M2.5 (minimaxm2.5) benchmark matrix entries from the active AMD and NVIDIA master configs so automated sweep generation no longer schedules them.

The deleted blocks are preserved verbatim in new archive files under .github/configs/deprecated/ (amd-minimaxm2.5-m2.7-master.yaml and nvidia-minimaxm2.5-m2.7-master.yaml), including fixed-seq-len, agentic-coding, atom/TRT, and multinode dynamo-vllm / vllm-disagg recipes. On NVIDIA, this also drops the large minimaxm2.5-fp4-gb300-dynamo-vllm section that previously sat ahead of the GLM-5 disagg entry (ordering-only cleanup in the active file).

^{Reviewed by Cursor Bugbot for commit 51c9441. Bugbot is set up for automated code reviews on this repo. Configure here.}

[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。