Add GLM-5.1 FP8 GB200 disaggregated dynamo-sglang MTP config / 添加 GLM-5.1 FP8 GB200 分离式 dynamo-sglang MTP 配置

[RohitNagraj]

Adds glm5-fp8-gb200-dynamo-sglang-mtp, a GB200 FP8 disaggregated
dynamo-sglang MTP config covering ISL/OSL 8k1k and 1k1k across
high-throughput (wide-EP) and low-latency (TP=8) decode topologies.

• Config points at the zip_override blocks of the vendored recipe at
  benchmarks/multi_node/srt-slurm-recipes/sglang/glm5/gb200-fp8/glm5-mtp.yaml,
  overlaid onto the srt-slurm checkout at launch.
• Chat template is enabled explicitly in the recipe
  (benchmark.use_chat_template: true) — MTP/EAGLE speculative decoding
  expects chat-formatted inputs.
• Image: lmsysorg/sglang:v0.5.13.post1-cu130.
• Adds the corresponding perf-changelog entry.

中文说明

新增 glm5-fp8-gb200-dynamo-sglang-mtp，一个 GB200 FP8 分离式（disaggregated）dynamo-sglang MTP 配置，覆盖 ISL/OSL 8k1k 与 1k1k 的高吞吐（wide-EP）和低延迟（TP=8）解码拓扑。

• 配置指向内置 recipe benchmarks/multi_node/srt-slurm-recipes/sglang/glm5/gb200-fp8/glm5-mtp.yaml 的 zip_override 块，启动时叠加到 srt-slurm 检出中。
• 在 recipe 中显式启用 chat template（benchmark.use_chat_template: true）—— MTP/EAGLE 投机解码要求 chat 格式的输入。
• 镜像：lmsysorg/sglang:v0.5.13.post1-cu130。
• 新增对应的 perf-changelog 条目。

[github-actions]

Thanks for the contribution! Please reach out to respective companies' CODEOWNER to fill in the latest PR_REVIEW_CHECKLIST.md before pinging core maintainer on Slack for review. In order for the signoff PR check bot to trigger, you must follow the PR_REVIEW_CHECKLIST.md template correctly, including the phrase As a PR reviewer and CODEOWNER, I have reviewed this and have.

For PR verification, add the full-sweep-fail-fast label (strongly recommended) to this PR — the benchmark sweep only runs on labeled PRs. Use full-sweep-enabled only if you need matrix jobs to keep running past a failure.

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. See GitHub's docs on re-running failed jobs

---

感谢你的贡献！请联系相应公司的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，然后再在 Slack 上联系核心维护者进行审阅。为了触发 signoff PR 检查机器人，你必须正确遵循 PR_REVIEW_CHECKLIST.md 模板，包括保留英文语句 As a PR reviewer and CODEOWNER, I have reviewed this and have。

如需进行 PR 验证，请为此 PR 添加 full-sweep-fail-fast 标签（强烈推荐）— 基准测试 sweep 仅在带有标签的 PR 上运行。仅当需要矩阵任务在失败后继续运行时才使用 full-sweep-enabled。

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。参见 GitHub 关于重新运行失败任务的文档

[github-actions]

Thanks for the contribution! Please reach out to respective companies' CODEOWNER to fill in the latest PR_REVIEW_CHECKLIST.md before pinging core maintainer on Slack for review. In order for the signoff PR check bot to trigger, you must follow the PR_REVIEW_CHECKLIST.md template correctly, including the phrase As a PR reviewer and CODEOWNER, I have reviewed this and have.

For PR verification, add the full-sweep-fail-fast label (strongly recommended) to this PR — the benchmark sweep only runs on labeled PRs. Use full-sweep-enabled only if you need matrix jobs to keep running past a failure.

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. See GitHub's docs on re-running failed jobs

---

感谢你的贡献！请联系相应公司的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，然后再在 Slack 上联系核心维护者进行审阅。为了触发 signoff PR 检查机器人，你必须正确遵循 PR_REVIEW_CHECKLIST.md 模板，包括保留英文语句 As a PR reviewer and CODEOWNER, I have reviewed this and have。

如需进行 PR 验证，请为此 PR 添加 full-sweep-fail-fast 标签（强烈推荐）— 基准测试 sweep 仅在带有标签的 PR 上运行。仅当需要矩阵任务在失败后继续运行时才使用 full-sweep-enabled。

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。参见 GitHub 关于重新运行失败任务的文档

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=28984702370
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=28984702370

[github-actions]

see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=28984702370
see unofficial run visualizer at https://inferencex.semianalysis.com/evaluation?unofficialRun=28984702370