[Klaud Cold] kimik2.5-fp4-b300-vllm: v0.26.0 image, extend conc to 512, add DEP4 arm / 升级 B300 镜像至 v0.26.0，扩展并发，新增 DEP4 - #2442
Conversation
|
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 For PR verification, add the 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 如需进行 PR 验证，请为此 PR 添加 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 |
xinli-sw
changed the title
xinli-sw
added
full-sweep-fail-fast
full-sweep-enabled
and removed
full-sweep-fail-fast
labels
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
|
|
||
| - config-keys: | ||
| - kimik2.5-fp4-b300-vllm | ||
| description: | ||
| - "Bump Kimi K2.5 NVFP4 B300 vLLM image from v0.22.0 to v0.26.0" | ||
| - "Extend TP8/TP4 concurrency sweep to 512" | ||
| - "Add DEP4 arm (tp4 ep4 dp-attn) with --prefill-schedule-interval 4, conc 128-512" | ||
| pr-link: TBD |
There was a problem hiding this comment.
🟡 The new perf-changelog.yaml entry for kimik2.5-fp4-b300-vllm sets pr-link: TBD instead of a real PR URL, breaking the changelog->PR traceability that every other entry in the file maintains. Since this is PR #2442, it should read pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2442.
Extended reasoning...
The new changelog block appended in this PR (perf-changelog.yaml, appended at the end of the file around line 5365) sets pr-link: TBD for the kimik2.5-fp4-b300-vllm entry:
- config-keys:
- kimik2.5-fp4-b300-vllm
description:
- "Bump Kimi K2.5 NVFP4 B300 vLLM image from v0.22.0 to v0.26.0"
- "Extend TP8/TP4 concurrency sweep to 512"
- "Add DEP4 arm (tp4 ep4 dp-attn) with --prefill-schedule-interval 4, conc 128-512"
pr-link: TBDEvery other entry in this 5300+ line file — including the immediately preceding kimik2.5-fp4-mi355x-vllm block a few lines above — carries a concrete https://github.com/SemiAnalysisAI/InferenceX/pull/NNNN link. A grep for pr-link: TBD across the file turns up exactly this one occurrence, confirming it's not an established convention but a placeholder that was never filled in before submission.
The purpose of pr-link in this changelog is to let anyone reading a config's history trace a given tuning change back to the PR that introduced it (including its description, discussion, and CI/benchmark results). TBD breaks that link for this entry specifically — a reader auditing why kimik2.5-fp4-b300-vllm was bumped to v0.26.0 or gained a DEP4 arm has no way to jump to the originating PR without external knowledge.
This is PR #2442 (per the PR metadata), so the fix is simply:
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2442Nothing breaks functionally if this is left as TBD — the changelog-trigger tooling fires on the entry's existence/content changing, not on the link being well-formed, and the benchmark sweep itself doesn't consume this field. So this is a documentation/traceability completeness gap rather than a functional bug: worth fixing before merge, but not blocking.
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30685318303 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30685980309 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30686074240 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30686418573 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30686474354 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30686474354 |
…-config to B300 script
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30698837294 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30698837294 |
…hdogs active across barriers) 中文：将 MiniMax-M3 AgentX 扫描固定到 AIPerf abf55f9（跨 barrier 保持空闲看门狗活跃） Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
…dle watchdogs active across barriers)" This reverts commit 367df34.
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30729674874 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30730026714 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30730985769 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30733192802 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30733234117 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30751776682 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=30755774107 |
1 participant
Summary
v0.22.0→vllm/vllm-openai:v0.26.0conc-end: 512tp: 4, ep: 4, dp-attn: true, conc-start: 128, conc-end: 512kimik2.5_fp4_b200.sh):DP_ATTENTIONhandling added — switches to--data-parallel-size, adds--enable-expert-paralleland--prefill-schedule-interval 4in DEP4 mode中文说明
v0.22.0→vllm/vllm-openai:v0.26.0conc-end: 512tp: 4, ep: 4, dp-attn: true, conc-start: 128, conc-end: 512kimik2.5_fp4_b200.sh）：新增DP_ATTENTION处理逻辑，DEP4 模式下切换为--data-parallel-size，并启用--enable-expert-parallel和--prefill-schedule-interval 4