Skip to content

[Klaud Cold] Update qwen3.5-fp4-b200-sglang-agentic-mtp SGLang image to nightly-dev-cu13-20260907-30705c00 #2862

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
functionstackx wants to merge 1 commit into
base: main
Choose a base branch
from
+9 −1
Open

[Klaud Cold] Update qwen3.5-fp4-b200-sglang-agentic-mtp SGLang image to nightly-dev-cu13-20260907-30705c00 #2862

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion configs/nvidia-master.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7548,7 +7548,7 @@ minimaxm3-fp8-h200-vllm-agentic-mtp:
- { tp: 8, spec-decoding: mtp, kv-offloading: dram, kv-offload-backend: { name: mooncake, version: "0.3.11.post1" }, conc-list: [12, 14] }

qwen3.5-fp4-b200-sglang-agentic-mtp:
image: lmsysorg/sglang:v0.5.16-cu130
image: lmsysorg/sglang:nightly-dev-cu13-20260907-30705c00
model: nvidia/Qwen3.5-397B-A17B-NVFP4
model-prefix: qwen3.5
runner: cluster:b200-nscale
Expand Down
8 changes: 8 additions & 0 deletions perf-changelog.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -6921,3 +6921,11 @@
- "Pick up the latest automatic ROCm DeepSeek-V4 optimizations, including fused mHC post/pre plus RMSNorm, gfx950 C4A top-k dispatch, fused C4 compressor GEMMs, fused SWA q/kv RMSNorm plus q FP8 quantization, and medium-batch cooperative top-k tuning."
- "Keep the existing VLLM_ROCM_USE_AITER=1, VLLM_ROCM_USE_AITER_MOE=1, and --moe-backend aiter settings, and explicitly add VLLM_ROCM_USE_AITER_FUSION_SHARED_EXPERTS=1 plus VLLM_ROCM_QUICK_REDUCE_QUANTIZATION=INT4 to both STP and MTP paths. The current checkpoint's shared-expert path does not satisfy the latest vLLM fusion conditions, so that fusion flag self-disables while preserving recipe parity."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2792

- config-keys:
- qwen3.5-fp4-b200-sglang-agentic-mtp
scenario-type:
- agentic-coding
description:

@claude claude Bot Sep 7, 2026

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.

🔴 The new changelog entry sets pr-link to the literal string "PRLINK_PLACEHOLDER", which is not the documented placeholder and will persist as garbage if the PR link is never manually fixed post-merge (docs/configuration-procedures.md:254 permits only TBD pre-merge, to be replaced with the real PR URL immediately after creation). Fix: use pr-link: TBD (or the actual PR URL once known) instead of a non-standard placeholder string, per the documented convention.

Extended reasoning...

docs/configuration-procedures.md line 254 states the only permitted placeholder before a PR exists is pr-link: TBD, replaced with the real URL right after PR creation. This diff instead writes pr-link: PRLINK_PLACEHOLDER (perf-changelog.yaml line 6929), a string that does not match any tooling or convention referenced in the docs. Since perf-changelog.yaml is append-only and byte-sensitive (never edited retroactively per AGENTS.md conventions), if this placeholder is not swept and replaced before/at merge, it becomes a permanent malformed record in the historical changelog, breaking any downstream tooling or scripts that parse pr-link fields expecting either a real GitHub URL or the TBD sentinel.

Verification: normal. perf-changelog.yaml:6929 sets pr-link: PRLINK_PLACEHOLDER on the newly appended entry. The repo's changelog gate rejects this value: utils/validate_perf_changelog.py:144-160 (validate_added_pr_link, invoked from compare_entries at line 206) requires an appended entry's link to be either the exact https://github.com/SemiAnalysisAI/InferenceX/pull/<pr_number> or a member of… | normal.…

- "Update SGLang image from lmsysorg/sglang:v0.5.16-cu130 (v0.5.16 release, cu130) to lmsysorg/sglang:nightly-dev-cu13-20260907-30705c00 (2026-09-07 cu13 dev nightly, digest sha256:19b8fa1223cc339c1eae7a5b703f1a8c2543b5b119155bf3d7efaef18f77f007, tag commit sgl-project/sglang@30705c00). Docker Hub last pushed the tag at 2026-09-07T01:43:42Z. Engine flags (modelopt_fp4, trtllm_mha, flashinfer_trtllm), NEXTN MTP settings, golden acceptance length 3.39, and the TP2/TP4 concurrency and HiCache grids are unchanged."
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/2862