Skip to content

[Klaud Cold] Update DSv4 FP4 MI355X SGLang AgentX MTP image to v0.5.21-rocm720-mi35x-20261010 / [Klaud Cold] 将 DSv4 FP4 MI355X SGLang AgentX MTP 镜像更新至 v0.5.21-rocm720-mi35x-20261010 - #3916

Open
Klaud-Cold wants to merge 2 commits into
mainfrom
klaud/auto-e73686e9e3949ff1-f9f80a5bd4994487
Open

Klaud-Cold wants to merge 2 commits into
mainfrom
klaud/auto-e73686e9e3949ff1-f9f80a5bd4994487

Conversation

@Klaud-Cold

@Klaud-Cold Klaud-Cold commented Oct 11, 2026 •

Copy link
Copy Markdown
Collaborator

Goal: Update SGLang image from lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261007 to lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673.
Baseline: 2026-10-10 · lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261007
Mean latency · Sources: API 1, API 2, API 3

Point Total tok/s/GPU ↑ Output tok/s/GPU ↑ TTFT ms ↓ TPOT ms ↓
AgentX c1 TP8 EP1 42c7f9 2,144.87 16.08 859.87 4.26
AgentX c4 TP8 EP1 60165f 3,444.63 25.46 625.02 4.63
AgentX c16 TP8 EP1 6d2bfa 11,411.4 78.8 1,055.87 7.29
AgentX c32 TP8 EP1 e249c3 21,640.8 137.42 1,721.39 10.28
AgentX c48 TP8 EP1 8a2631 28,267.65 193.62 2,548.78 12.4
AgentX c128 TP8 EP1 c19c1d 41,998.11 365.76 7,580.58 27.84
AgentX c128 TP8 EP8 5c3e5b 42,755.68 372.8 7,002.87 27.54
AgentX c256 TP8 EP8 e7745a 67,295.19 514.89 20,801.97 32.52
AgentX c384 TP8 EP8 6f09ec 74,092.86 578.85 39,957.13 40.74
Eval Score ↑ Samples
gsm8k/em_strict · c48 TP8 EP1 97.57% 1,319
gsm8k/em_strict · c384 TP8 EP8 DPA 97.73% 1,319
中文

**目标:**将 SGLang 镜像从 lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261007 更新为 lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673。
**基线:**2026-10-10 · lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261007
平均延迟 · 来源: API 1, API 2, API 3;数值及异常说明见上表。

…5x-20261010

Pin lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010 by digest in the
master config and its srt-slurm recipe.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@github-actions

Copy link
Copy Markdown
Contributor

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

Klaud-Cold commented Oct 11, 2026 •

Copy link
Copy Markdown
Collaborator Author

Initial attempt · Passed · Run 38141232733 / attempt 1 · 2026-10-11 15:07 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · c801c9f2dbbd · Mean latency
Change: Pin the 20261010 SGLang ROCm nightly by digest; it bundles SGLang a916d91cd3 instead of aa5551d9b6 with unchanged AITER, Triton, MoRI and Mooncake commits, and every recipe flag and environment variable still exists at the new commit.

Point Output tok/s/GPU ↑ TTFT ms ↓ TPOT ms ↓
AgentX c1 TP8 EP1 42c7f9 16.09 (N/A) 843.12 (N/A) 4.21 (N/A)
AgentX c32 TP8 EP1 e249c3 138.11 (N/A) 1,617.62 (N/A) 10.03 (N/A)
AgentX c128 TP8 EP1 c19c1d 370.58 (N/A) 7,181.25 (N/A) 27.42 (N/A)
AgentX c128 TP8 EP8 5c3e5b 371.3 (N/A) 7,118.91 (N/A) 27.45 (N/A)

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

Eval Score ↑ Samples
gsm8k/em_strict · c48 TP8 EP1 96.97% (-0.61 pp) 1,319 each
gsm8k/em_strict · c384 TP8 EP8 DPA 97.12% (-0.61 pp) 1,319 each

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

中文

初次尝试 · 已通过 · Run 38141232733 / attempt 1 · 2026-10-11 15:07 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · c801c9f2dbbd · 平均延迟
**变更:**按摘要固定 20261010 SGLang ROCm nightly 镜像;其内置 SGLang 从 aa5551d9b6 更新为 a916d91cd3,AITER、Triton、MoRI 与 Mooncake 提交不变,配方使用的参数和环境变量在新提交中仍然存在。实测数值及异常说明见上表。
**下一步:**追加 perf-changelog 条目并运行最终完整 sweep。

…update

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@Klaud-Cold Klaud-Cold added the full-sweep-fail-fast Full sweep with canary gate; first failure cancels the rest of that matrix (recommended) label Oct 11, 2026
@Klaud-Cold

Klaud-Cold commented Oct 11, 2026 •

Copy link
Copy Markdown
Collaborator Author

Final full sweep · Running · Run 38149983765 / attempt 1 · 2026-10-11 17:39 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · 60f3e2f1c50a · Mean latency
Change: Add the perf-changelog entry for the smoke-validated image and run the full sweep.

Point Output tok/s/GPU ↑ TTFT ms ↓ TPOT ms ↓
AgentX c1 TP8 EP1 42c7f9 N/A N/A N/A
AgentX c4 TP8 EP1 60165f N/A N/A N/A
AgentX c16 TP8 EP1 6d2bfa N/A N/A N/A
AgentX c32 TP8 EP1 e249c3 N/A N/A N/A
AgentX c48 TP8 EP1 8a2631 N/A N/A N/A
AgentX c128 TP8 EP1 c19c1d N/A N/A N/A
AgentX c128 TP8 EP8 5c3e5b N/A N/A N/A
AgentX c256 TP8 EP8 e7745a N/A N/A N/A
AgentX c384 TP8 EP8 6f09ec N/A N/A N/A

Note: AgentX c1 TP8 EP1 42c7f9, AgentX c16 TP8 EP1 6d2bfa, AgentX c32 TP8 EP1 e249c3, AgentX c48 TP8 EP1 8a2631, AgentX c128 TP8 EP8 5c3e5b: cancelled.
Note: All rows: request errors unavailable.
Note: All rows: Δ N/A: point unavailable.
Note: AgentX c4 TP8 EP1 60165f: failed.
Note: AgentX c128 TP8 EP1 c19c1d, AgentX c256 TP8 EP8 e7745a, AgentX c384 TP8 EP8 6f09ec: pending.

Eval Score ↑ Samples
gsm8k/em_strict · c48 TP8 EP1 N/A 1,319/N/A (old/new)
gsm8k/em_strict · c384 TP8 EP8 DPA N/A 1,319/N/A (old/new)

Note: All rows: unavailable.
Note: All rows: Δ N/A: eval score unavailable.

Next: Wait for the running jobs, then rerun the failed and cancelled jobs on the same head as an infrastructure retry: node mia1-p01-g15 still held memory from a prior job and its GPU drain gate timed out before SGLang started.

中文

最终完整 sweep · 运行中 · Run 38149983765 / attempt 1 · 2026-10-11 17:39 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · 60f3e2f1c50a · 平均延迟
**变更:**为冒烟验证通过的镜像追加 perf-changelog 条目并运行完整 sweep。实测数值及异常说明见上表。
**下一步:**等待运行中的作业结束后,在同一 head 上重跑失败和被取消的作业(基础设施重试):节点 mia1-p01-g15 仍有上一个作业的显存占用,GPU 释放检查在 SGLang 启动前超时。

@github-actions

Copy link
Copy Markdown
Contributor

@Klaud-Cold

Klaud-Cold commented Oct 11, 2026 •

Copy link
Copy Markdown
Collaborator Author

Final full sweep · Passed · Run 38149983765 / attempt 2 · 2026-10-12 00:32 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · 60f3e2f1c50a · Mean latency
Change: Validate the complete updated-image family at this head.

Point Output tok/s/GPU ↑ TTFT ms ↓ TPOT ms ↓
AgentX c1 TP8 EP1 42c7f9 15.99 (N/A) 878.07 (N/A) 4.27 (N/A)
AgentX c4 TP8 EP1 60165f 26.17 (N/A) 664.9 (N/A) 4.6 (N/A)
AgentX c16 TP8 EP1 6d2bfa 79.06 (N/A) 1,044.63 (N/A) 7.2 (N/A)
AgentX c32 TP8 EP1 e249c3 138.28 (N/A) 1,580.88 (N/A) 10 (N/A)
AgentX c48 TP8 EP1 8a2631 193.87 (N/A) 2,430.66 (N/A) 12.49 (N/A)
AgentX c128 TP8 EP1 c19c1d 366.72 (N/A) 7,388.97 (N/A) 28.1 (N/A)
AgentX c128 TP8 EP8 5c3e5b 368.37 (N/A) 7,362.49 (N/A) 27.76 (N/A)
AgentX c256 TP8 EP8 e7745a 519.26 (N/A) 20,362.92 (N/A) 32.52 (N/A)
AgentX c384 TP8 EP8 6f09ec 596.65 (N/A) 36,908.96 (N/A) 40.69 (N/A)

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

Eval Score ↑ Samples
gsm8k/em_strict · c48 97.42% (-0.15 pp) 1,319 each
gsm8k/em_strict · c384 97.12% (-0.61 pp) 1,319 each

Next: Mark ready for maintainer review.

中文

最终完整 sweep · 已通过 · Run 38149983765 / attempt 2 · 2026-10-12 00:32 UTC
lmsysorg/sglang-rocm:v0.5.21-rocm720-mi35x-20261010@sha256:abc9e62d3f2b6f9dd3284619900cef7066a60552d0440e36c608801e0aeb5673 · 60f3e2f1c50a · 平均延迟
**变更:**验证此提交更新镜像后的完整配置族。实测数值及异常说明见上表。
**下一步:**标记为就绪,等待维护者审查。

@Klaud-Cold
Klaud-Cold marked this pull request as ready for review October 12, 2026 00:32
@Klaud-Cold
Klaud-Cold requested a review from a team October 12, 2026 00:32
@Klaud-Cold

Copy link
Copy Markdown
Collaborator Author

validated · Repairs: unknown · Runs: 38141232733, 38149964881, 38149983765
All owned runs ended. Full sweep verified; ready for review.

中文

validated · 修复次数:unknown · 运行:38141232733, 38149964881, 38149983765
所有自有运行均已结束。完整 sweep 已验证;已就绪,等待审查。

@claude

claude Bot commented Oct 12, 2026 •

Copy link
Copy Markdown
Contributor

Claude finished @Klaud-Cold's task in 17s —— View job


LGTM - no blocking issues found

@claude claude Bot left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

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

Looks good, straightforward config change.

What was reviewed: the image tag+digest bump in agentic.yaml's model.container and amd-master.yaml's image — both updated identically as required; the perf-changelog.yaml diff — confirmed it only appends one new English-only block at the tail referencing PR #3916, with no existing bytes altered.

Extended reasoning...

Three-file diff: a pinned SGLang ROCm container image bump (tag + new sha256 digest) applied identically to the srt-slurm recipe and the matching amd-master.yaml entry, plus one append-only perf-changelog.yaml entry. No code, auth, or security-sensitive paths are touched — purely declarative YAML config. The change is small, mechanical, internally consistent, and follows the repo's documented conventions for recipe/master-config sync and changelog append-only format, so human review is not essential.

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

full-sweep-fail-fast Full sweep with canary gate; first failure cancels the rest of that matrix (recommended)

Projects

Status: No status

Development

Successfully merging this pull request may close these issues.

1 participant