[AMD][MI355X] DSv4 SGLang agentic: v0.5.14 image, router, HiCache#2146
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 |
seungrokj
added
AMD
full-sweep-fail-fast
agentx
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
| *) | ||
| echo "Error: unsupported KV_OFFLOAD_BACKEND '${KV_OFFLOAD_BACKEND:-}' (expected: native, mooncake, lmcache)" >&2 | ||
| exit 1 |
There was a problem hiding this comment.
🟡 The default arm of the KV_OFFLOAD_BACKEND case at line 85 prints (expected: native, mooncake, lmcache), but the case above only accepts hicache — the listed values look copy-pasted from the sibling vLLM launchers (which do accept native/mooncake/lmcache) and would themselves fall into the same default arm here. Sweep configs only use hicache so CI never trips this branch, but anyone debugging by hand will be misdirected. Fix: change the list to (expected: hicache).
Extended reasoning...
What the bug is
The case "${KV_OFFLOAD_BACKEND:-}" block at lines 56-88 of benchmarks/single_node/agentic/dsv4_fp4_mi355x_sglang.sh has exactly one accepting arm: hicache). Anything else falls into the default *) arm at lines 84-87, which emits:
Error: unsupported KV_OFFLOAD_BACKEND '<value>' (expected: native, mooncake, lmcache)
None of native, mooncake, or lmcache are recognized by this case statement — they are all backends supported by other launchers in the repo (native for kimik2.5-fp4-mi355x-vllm-agentic, mooncake for minimaxm3-fp8-mi300x-vllm-agentic, lmcache for the sibling dsv4-fp4-mi355x-vllm-agentic). The message is a copy-paste from those vLLM-family launchers.
Why the message is self-contradictory
A user who sees unsupported KV_OFFLOAD_BACKEND 'foo' and reads the parenthetical hint will pick one of native, mooncake, or lmcache — and hit the exact same default arm and the exact same message. In the case where they picked native, the output literally reads unsupported KV_OFFLOAD_BACKEND 'native' (expected: native, mooncake, lmcache), which is self-contradictory.
Step-by-step proof
- Operator invokes the launcher manually with
KV_OFFLOADING=dram KV_OFFLOAD_BACKEND=native(a natural attempt after seeing another dsv4 launcher uselmcache). agentic_kv_offload_enabledreturns true becauseKV_OFFLOADING=dram.- Bash enters the
caseat line 56 and evaluates the arms in order. The only arm ishicache). nativedoes not matchhicache), so control falls to*)at line 84.- Line 85 prints:
Error: unsupported KV_OFFLOAD_BACKEND 'native' (expected: native, mooncake, lmcache). - Line 86 exits 1.
- Operator, taking the message at face value, retries with
mooncakeorlmcache. Same path — steps 3-6 repeat.
Why existing code doesn't prevent it
The header comment at line 7 states "KV_OFFLOADING=dram requires KV_OFFLOAD_BACKEND=hicache", so the canonical documented value is hicache. The error text just wasn't updated to match when this launcher was adapted from the vLLM sibling.
Impact
Cosmetic misdirection in a diagnostic-only path. The four sweep entries in configs/amd-master.yaml for this recipe all pass kv-offload-backend: hicache, so CI never hits the default arm. The cost is a minute or two of wasted debug time for someone experimenting with a non-hicache backend by hand.
Fix
Change line 85's parenthetical to (expected: hicache):
echo "Error: unsupported KV_OFFLOAD_BACKEND '${KV_OFFLOAD_BACKEND:-}' (expected: hicache)" >&2On the refutations
Both refutations are duplicate-management: bug_002 and bug_004 target the same defect at the same lines with the same fix, and the refuters explicitly note the finding itself is real — they only refuted one of the two to avoid posting the same comment twice. The synthesis agent merged them into a single report, which resolves that concern.
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29076727890 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29077198891 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29081203249 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29081361033 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29094196523 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29108708524 |
functionstackx
left a comment
functionstackx
left a comment
There was a problem hiding this comment.
please update sglang cookbook with ur current optimized recipes, sglang cookbook has support for HiCache
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29108764431 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29108764431 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29136063462 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29136075686 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29136311108 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29136685794 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29136685794 |
|
@functionstackx can you plz review this? |
would u like to address my last comment? #2146 (review) |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29379931245 |
1 similar comment
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29379931245 |
…ache sweep to c96 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29393967098 |
…t HiCache sweep to c64 Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29401128784 |
…ore full sweep entries Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
seungrokj
changed the title
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29407950932 |
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29408074234 |
|
@functionstackx @cquil11 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29413860950 |
2 similar comments
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29413860950 |
|
see unofficial run visualizer at https://inferencex.semianalysis.com/inference?unofficialRun=29413860950 |
|
/reuse-sweep-run |
|
/stage-results 29413860950 |
|
@cquil11 staged run 29413860950: https://inferencemax-app-git-staging-semianalysisai.vercel.app/inference?i_dates=2026-07-15~r29413860950 This shared staging slot remains available until the next |
There was a problem hiding this comment.
As a PR reviewer and CODEOWNER, I have reviewed this and have:
- Verified that as of the moment of typing this, this is the latest version of PR_REVIEW_CHECKLIST.md
- Verified that the general code quality meets the InferenceX standard and does not make the code quality any worse.
- Verified that this PR has passed PR validation. Please link to GitHub Action workflow that shows this. Link: https://github.com/SemiAnalysisAI/InferenceX/actions/runs/29413860950
- Verified that this PR passes evals. Please link to GitHub Action workflow that shows this. Eval currently missing from https://github.com/SemiAnalysisAI/InferenceX/actions/runs/29413860950 as we haven't enable this yet in AgentX.
- Verified that speculative decoding PRs uses chat templates to align the AL distribution to real world
- For agentic workloads: verified that speculative-decoding configs (EAGLE / MTP / draft models) run with simulated synthetic acceptance, with the acceptance-length value taken from the committed golden AL curve in golden_al_distribution/ for that model, thinking mode, and draft length. A submission may choose any supported draft length, but it may not substitute a different acceptance target.
- Verified that the model architecture isn't changed with benchmark hacks like using --hf-overrides to skipping indexer for every x layers on models that don't natively support this. As a general rule, we won't accept optimizations that reduces the number of model architecture FLOPs. Anything that makes that same computation run faster is fair game; FLOPs at lower precisions is fine, given that the config passes private evals. As an general north star princple, we should only use optimizations which is used in production by customers that care about accuracy
- If an company claims that they support vLLM/SGLang as first class LLM inference engines on their hardware, I have verified that the respective vLLM submission made using upstream https://hub.docker.com/u/vllm docker repo, upstream SGLang https://hub.docker.com/u/lmsysorg docker repo. The only exceptions are for new hardware, such as MI455X UALoE72, Vera Rubin NVL72, Rubin NVL8, etc., and for new model architectures where there is an actual reason why vLLM/SGLang does not fundamentally support them yet as supported by vLLM/SGLang community maintainers
- If an company claims that they support vLLM/SGLang as first class upstream in-tree LLM inference engines on their hardware, I have have verified that the respective vLLM/SGLang submission has been made before additional frameworks (TRT-LLM, ATOM, etc.). The only exceptions are for new hardware, such as MI455X UALoE72, Vera Rubin NVL72, Rubin NVL8, etc., and for new model architectures where there is an actual reason why vLLM/SGLang does not fundamentally support them yet.
- Verified that every single-node vLLM/SGLang recipe in this PR is documented in the official vLLM recipes and/or the SGLang cookbook:
- I linked the corresponding upstream PR in the vLLM recipe repo or SGLang repo and verified that it is MERGED before this InferenceX PR merges. An opened, draft, or closed-without-merge upstream PR does not satisfy this requirement. If the matching recipe was already published, I linked the published recipe/cookbook page in the additional detail section below.
- Verified that this PR does not patch the inference engine or serving stack — the pinned image must run as shipped. This covers .patch files / git apply / patch, inline patches embedded in benchmark scripts (e.g. a python3/sed heredoc that rewrites installed engine sources before serving), in-place edits of site-packages, monkey-patching, overwriting container files, and installing forked/rebuilt engine wheels on top of the pinned image. The only exception is a patch covered by a filled-out waiver at docs/waiver/
<PR_NUMBER>.md— named after the PR that introduces the patch and filed in that same PR, stating what is patched, why the unmodified upstream image cannot run this benchmark, the upstream PR/issue link, and the removal plan — which I have linked below in the additional detail section. - If any of the above criteria cannot reasonably be satisfied, I have provided additional reasoning below.
Additional detail section:
- insert any additional info here
Cookbook PR (merged): sgl-project/sglang#31122
Signed: @chunfangamd
4 participants
Summary
dsv4-fp4-mi355x-sglang-agenticimage tolmsysorg/sglang-rocm:v0.5.14-rocm720-mi35x-20260710dsv4attention backend,fp8_e4m3KV cache,--disable-radix-cache,--cuda-graph-max-bs, DP-attention exports,--enable-two-batch-overlap)SGLANG_CMDarray pattern with command logging and env-var dumpcapture_cache_metricsfor pre/post-benchmark cache statsconc=48across TP8 +/- DPA +/- HiCacheTest plan
dsv4-fp4-mi355x-sglang-agenticfull sweep on MI355X cluster🤖 Generated with Claude Code