Commits vllm-project/vllm

Merge branch 'docs/model-recipes-skill' of github.com:esmeetu/vllm into docs/model-recipes-skill

esmeetu committed 2 days ago

Verified a06debb2

Expose model recipes skill to Claude Code

esmeetu committed 2 days ago

Verified 235f16f3

Merge branch 'main' into docs/model-recipes-skill

esmeetu committed 2 days ago

Verified 1fee2b7d

[Bugfix][Model] Fix M-RoPE offset double-count in Qwen3-Omni (#58890)

C1ves committed 2 days ago

Verified 44198f57

[TEST][XPU][CI] disable xpu tests for nonexistent input norm kernels (#59568)

microslaw committed 2 days ago

Verified 8c3b720a

[EPLB] Add contention-aware expert migration batching (#52641)

DOCCA0 committed 2 days ago

Verified 7dfe3338

[CI/Build][NVIDIA] Build Rubin images on the public nvidia/cuda base image (#59288)

wangshangsam committed 2 days ago

Verified 8e443871

[GLM 5.3 Perf] Enable fused multi-step decode, 13.3% E2E throughput improvement for concurrency 1 (#57443)

yewentao256 committed 2 days ago

Verified 8bdd8d84

[Docs] Add ERNIE 4.5 to batch invariance tested models (#58476)

yifanFengg committed 2 days ago

Verified 6e517b15

[Perf][MoE] Support fp8 combine in FlashInfer one-sided MoE all2all (#57995)

Ayu190505 committed 2 days ago

Verified 84593955

[feat] FlashInfer CuteDSL MegaMoE integration (#54049)

jdebache committed 2 days ago

Verified 111f71a6

[Bugfix][Frontend] Avoid generation for empty streaming input (#59015)

nvbfalk committed 2 days ago

Verified c4973d94

[Perf] Use value-only reduction for native per-token FP8 quantization (#59800)

jackLei0901 committed 2 days ago

Verified 10f6ab01

[Bugfix] Bump tokenizers to 0.23.2 for duplicate-pattern support (#59796)

djramic committed 2 days ago

Verified 8882b83b

[Bugfix][Watermarking] Keep draft prompt lengths valid under CUDA graphs (#59779)

simon-veitner-redhat committed 2 days ago

Verified 202d497a

[Frontend] Constrain non-strict GLM-4.7 tool calls with a shallow structural tag (#56403)

Dovis01 committed 2 days ago

Verified d1a974c1

[Refactor] Remove dead env and config (#59781)

yewentao256 committed 2 days ago

Verified f700178f

[GLM5.3 Perf] Reuse sparse MLA index conversion across layers, 3.5~3.9x kernel performance improvement (#59464)

yewentao256 committed 2 days ago

Verified 4e4d75c4

Revert "[Bugfix] Tie lm_head.weight for Nemotron Parse when checkpoint omits it" (#53020) (#59805)

Thangnguyenvn98 committed 2 days ago

Verified 7cfd2335

[Perf][Qwen4Exp] Add SM121 TP=1 skinny-GEMM plans (#59753)

stecasta committed 2 days ago

Verified f590eb24

[Perf] Tune MoE weighted-sum kernel launch configuration (#59731)

jinzhen-lin committed 2 days ago

Verified 9cf087fd

Upgrade tpu-inference to v0.30.0 (#59396)

Jeff-Tseng-dev committed 2 days ago

Verified 84421174

[Agents] Add pre-commit check for agent files and skills (#59657)

mgoin committed 2 days ago

Verified 1ab18599

[Minimax-M3] Keep the native FP8 MMA in the Triton indexer scorers (#59481)

rmhaskarnvidia committed 2 days ago

Verified e42d35d4

[CI] Fix ModelExpress handling in weight transfer tests (#59772)

djramic committed 2 days ago

Verified 167a81fa

[ROCm][CI] Add test coverage for VLLM_ROCM_MOE_PADDING memory-stride padding transparency (#59332)

divakar-amd committed 2 days ago

Verified 6abadc2a

[CI] Use vllm_runner in fusions_e2e conftest for reliable GPU cleanup (#59525)

divakar-amd committed 2 days ago

Verified 2a537887

[Docs] Add Reviewers page (#59459)

LucasWilkinson committed 2 days ago

Verified 7b6be2f1

[Bugfix] Bind routed-experts capture to the MoE layer, not the kernel (#59455)

erictang000 committed 2 days ago

Verified 4c4003bb

[Bugfix][Multimodal] Fix GLM-5.3-Flash vision tower crashes on image input (#59126)

gaby committed 2 days ago

Verified 09798929