[Klaud-Cold]

Goal: Update vLLM image from vllm/vllm-openai:nightly-dev-x86_64-cu130-ac9126e58aa7 to vllm/vllm-openai:nightly-5555a2d8f7116b2b4877f66516f4128176cf722a@sha256:8896fd7b29fc4ee7ce839f146cb4757984c476ec79d0234ffa8c995dee6e339a.
Baseline: 2026-10-02 · vllm/vllm-openai:nightly-dev-x86_64-cu130-ac9126e58aa7
Mean latency · Sources: API 1, API 2, API 3, API 4

Point	Total tok/s/GPU ↑	Output tok/s/GPU ↑	TTFT ms ↓	TPOT ms ↓
AgentX c1 TP4 EP1 f41bbd	7,537.41	45.46	316.92	1.69
AgentX c4 TP4 EP1 a60967	10,268.87	70.09	211.63	1.82
AgentX c8 TP2 EP2 c75e8c	30,793.29	222.51	213.18	2.92
AgentX c8 TP4 EP1 74519e	16,585.72	115.03	175.2	2.04
AgentX c16 TP2 EP2 c95c57	56,827.42	358.48	200.7	3.38
AgentX c16 TP4 EP1 5f4875	29,331.81	190.22	187.76	2.56
AgentX c32 TP2 EP2 2e8842	119,446.55	687.84	221.67	4.64
AgentX c64 TP2 EP2 7b01d6	214,245.91	1,525.81	268.59	7.81
AgentX c128 TP2 EP2 0e11a9	299,060.19	2,406.26	389.92	12.95
AgentX c192 TP2 EP2 b6fb07	406,373.87	3,059.33	513.06	18.05

Eval	Score ↑	Samples
gsm8k/em_strict · c16 TP4 EP1	97.35%	1,319
gsm8k/em_strict · c192 TP2 EP2	97.27%	1,319

中文

**目标：**将 vLLM 镜像从 vllm/vllm-openai:nightly-dev-x86_64-cu130-ac9126e58aa7 更新为 vllm/vllm-openai:nightly-5555a2d8f7116b2b4877f66516f4128176cf722a@sha256:8896fd7b29fc4ee7ce839f146cb4757984c476ec79d0234ffa8c995dee6e339a。
**基线：**2026-10-02 · vllm/vllm-openai:nightly-dev-x86_64-cu130-ac9126e58aa7
平均延迟 · 来源： API 1, API 2, API 3, API 4；数值及异常说明见上表。

[github-actions]

Thanks for the contribution!

• Review: If this PR changes files owned by someone other than a repository admin or @SemiAnalysisAI/core, ask one eligible CODEOWNER to complete the latest PR_REVIEW_CHECKLIST.md before contacting a core maintainer on Slack. Follow the template exactly, including As a PR reviewer and CODEOWNER, I have reviewed this and have, so sign-off verification triggers.
• PR verification: Sweeps run only when the PR appends an inferencex-e2e/perf-changelog.yaml entry and carries exactly one primary label: full-sweep-fail-fast (strongly recommended; canary plus per-matrix fail-fast), full-sweep-enabled (canary; matrix jobs continue after a failure), or non-canary-full-sweep-enabled (no canary or fail-fast). The modifiers all-evals, evals-only, and agentx-fast require a primary label. On fork PRs, a maintainer applies the label. See sweep labels and reuse.
• After merging: PR authors must ensure all GitHub Actions jobs pass. Transient failures often pass on rerun; see how to rerun failed jobs.

中文

感谢你的贡献！

• **审阅：**如果 PR 修改的文件归属于仓库管理员及 @SemiAnalysisAI/core 之外的 CODEOWNER，请先联系一位有资格的 CODEOWNER 填写最新的 PR_REVIEW_CHECKLIST.md，再通过 Slack 联系核心维护者。必须严格遵循模板，并保留 As a PR reviewer and CODEOWNER, I have reviewed this and have，才能触发签核验证。
• **PR 验证：**只有当 PR 在 inferencex-e2e/perf-changelog.yaml 末尾追加了条目，并且恰好带有一个主标签时，才会运行扫描：full-sweep-fail-fast（强烈推荐；canary 加逐矩阵 fail-fast）、full-sweep-enabled（有 canary；矩阵任务在失败后继续运行）或 non-canary-full-sweep-enabled（无 canary，也无 fail-fast）。修饰标签 all-evals、evals-only 和 agentx-fast 必须与主标签一起使用。fork PR 的标签由维护者添加。参见扫描标签与复用。
• **合并后：**PR 作者必须确保所有 GitHub Actions 任务通过。临时性失败通常可以通过重新运行恢复；参见重新运行失败任务的说明。

[Klaud-Cold]

Initial attempt · Passed · Run 38360124060 / attempt 1 · 2026-10-13 09:21 UTC
vllm/vllm-openai:nightly-5555a2d8f7116b2b4877f66516f4128176cf722a@sha256:8896fd7b29fc4ee7ce839f146cb4757984c476ec79d0234ffa8c995dee6e339a · f85e1b9a259b · Mean latency
Change: Measured PR head 6ee068eb: swap the image to vLLM main nightly 5555a2d8 (635 commits after ac9126e5, same CUDA 13.0.2 pipeline), recipe flags unchanged: every used option still parses, and the new source adds native decoder_replay_cudagraph_capture_sizes (#59532) within max-num-batched-tokens; the old image's labels claim ac9126e5, which lacks that field, so old provenance is uncertain.

Point	Output tok/s/GPU ↑	TTFT ms ↓	TPOT ms ↓
AgentX c1 TP4 EP1 f41bbd	45.97 (N/A)	312.42 (N/A)	1.62 (N/A)
AgentX c8 TP2 EP2 c75e8c	223.38 (N/A)	153.61 (N/A)	2.91 (N/A)

Note: All rows: Δ N/A: dataset/scenario/statistic mismatch.

Eval	Score ↑	Samples
gsm8k/em_strict · c16 TP4 EP1	97.57% (+0.23 pp)	1,319 each
gsm8k/em_strict · c192 TP2 EP2	97.27% (+0.00 pp)	1,319 each

Next: Append the changelog entry and run the final full sweep.

中文

初次尝试 · 已通过 · Run 38360124060 / attempt 1 · 2026-10-13 09:21 UTC
vllm/vllm-openai:nightly-5555a2d8f7116b2b4877f66516f4128176cf722a@sha256:8896fd7b29fc4ee7ce839f146cb4757984c476ec79d0234ffa8c995dee6e339a · f85e1b9a259b · 平均延迟
**变更：**测量的 PR head 为 6ee068eb：将镜像更换为 vLLM main nightly 5555a2d8（比 ac9126e5 新 635 个提交，同为 CUDA 13.0.2 构建流程），配方参数不变：所用选项均仍可解析，新源码原生支持 decoder_replay_cudagraph_capture_sizes（#59532），且不超过 max-num-batched-tokens；旧镜像标签声明为 ac9126e5，但该提交不含此字段，旧镜像来源存在不确定性。实测数值及异常说明见上表。
**下一步：**追加 changelog 条目并运行最终完整 sweep。

[Klaud-Cold]

capacity-deferred · Cleanup pending. Stop and confirm owned runs before closing.

中文

capacity-deferred · 清理待完成。先停止并确认自有运行结束，再关闭 PR。

[Klaud-Cold]

capacity-deferred · Repairs: 0 · Runs: 38360124060, 38375737313
All owned runs ended. PR closed; branch deleted for retry.

中文

capacity-deferred · 修复次数：0 · 运行：38360124060, 38375737313
所有自有运行均已结束。PR 已关闭；分支已删除，可重新尝试。