Skip to content

Commit 562aa3f

Browse files
committed
Append perf-changelog entry for PR #1586
1 parent 43bae76 commit 562aa3f

1 file changed

Lines changed: 8 additions & 0 deletions

File tree

perf-changelog.yaml

Lines changed: 8 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -3395,3 +3395,11 @@
33953395
description:
33963396
- "Add DeepSeek-V4-Pro FP4 MI355X ATOM MTP3 benchmark; image rocm/atom:rocm7.2.4_ubuntu24.04_py3.12_pytorch_release_2.10.0_atom0.1.3"
33973397
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1627
3398+
3399+
- config-keys:
3400+
- dsv4-fp4-gb300-dynamo-sglang
3401+
description:
3402+
- "Add wide-EP sweep configs (EP=12/16/24/32/40) matching srt-slurm PR#173 topology (18 nodes total)"
3403+
- "EP=12 15P+3D conc=12000, EP=16 14P+4D conc=8192, EP=24 12P+6D conc=3000, EP=32 10P+8D conc=2500, EP=40 8P+10D conc=2048"
3404+
- "Aligned decode params with Weiliang config: swa-full-tokens-ratio=0.20, max-running-requests=18432, moe-dense-tp-size=1; added prefill enable-dp-lm-head and cuda-graph-max-bs=512"
3405+
pr-link: https://github.com/SemiAnalysisAI/InferenceX/pull/1586

0 commit comments

Comments
 (0)