vllm-project/vllm

Pull Requests Commits

Merge branch 'main' into fi-ar-pdl

jeejeelee committed 3 hours ago

Verified cd4533db

[CI] Add DSV4-Flash to gsm8k moe-refactor/config-b200.txt (#42111)

mgoin committed 3 hours ago

Verified cd0ff26e

jeejeelee committed 3 hours ago

ad2465b0

[Perf] Avoid forward scan for async output placeholders (#42938)

izikgo committed 3 hours ago

Verified 2ae910ed

add enqueue all option to throughput benchmark (#42975)

pmaybank committed 3 hours ago

Verified fadf5d33

[Perf][Bugfix] Update dflash aux layer indexing (#40727)

benchislett committed 3 hours ago

Verified c628a93a

[Cohere] Enable Cohere MoE (#43143)

Terrencezzj committed 4 hours ago

Verified 5774aaed

[MRV2][BugFix] Fix default-stream CG capture in P/W LoRA case (#43160)

njhill committed 4 hours ago

Verified 39bba710

[bug] fix WeightTransferConfig.backend to allow for all strings (#43121)

hao-aaron committed 5 hours ago

Verified 73dd2f33

[CPU][DOC] Fix installation commands for Arm CPUs (#43115)

fadara01 committed 7 hours ago

Verified be167859

Fix error in Dynamic NTK scaling (#41277)

maxdebayser committed 9 hours ago

Verified 117afeea

[Model] Support post-norm architecture for EAGLE-3 supeculators (#42764)

Dogacel committed 10 hours ago

Verified 12421962

[ci] Move language models tests (hybrid) back to L4 (#43129)

khluu committed 11 hours ago

Verified a65093c1

[CI failure] Temporarily disable using persistent cache for flashinfer autotune (#43119)

wzhao18 committed 12 hours ago

Verified 9aaf83ef

[Bugfix][MoE] FlashInfer one-sided: workspace union across heterogeneous layers (#42976)

tomeras91 committed 12 hours ago

Verified f54721bc

[Docs] Fix MooncakeStoreConnector role in disaggregated example (#42994)

Dao007forever committed 12 hours ago

Verified aed2eb35

[feat] Add FP8 per-tensor Q scale support to Triton attention backend (#42080)

DomBrown committed 14 hours ago

Verified d247a931

[Misc] add humming to dependencies (#42540)

jinzhen-lin committed 15 hours ago

Verified 8200fbe1

[Refactor] Extract extract_types_from_schema utility from Minimax M2 tool parser (#43025)

sfeng33 committed 15 hours ago

Verified 42b4f1fd

[Model] Openvla support (#42654)

yiwen101 committed 15 hours ago

Verified 1c615808

[XPU] update xpu graph usage (#43043)

xinyu-intel committed 15 hours ago

Verified d740e2c0

[Perf][4/n] Eliminate various GPU<->CPU syncs (#42347)

njhill committed 16 hours ago

Verified b82e908b

jeejeelee committed 17 hours ago

79866717

[Bugfix] Fix top logprobs token placeholders in `/inference/v1/generate` (#42887)

sagearc committed 20 hours ago

Verified a78b842d

[CI] Add MTP + PD disagg test for Qwen3.5 (#42677)

ZhanqiuHu committed 21 hours ago

Verified 129019f3

[Misc][MM] Remove redundant code in CLIPAttention (#43046)

shen-shanshan committed 22 hours ago

Verified ef54a4d6

[Model Refactoring] Rename deepseek_v4.py to model.py [4/N] (#43077)

WoosukKwon committed 22 hours ago

Verified 07beaed8

[KVConnector][DSV4] HMA support for Mooncake store connector (#42828)

ivanium committed 22 hours ago

Verified 056bc2e1

[bug] AsyncScheduler drops first post-resume token after pause_generation + clear_cache (#42117)

hao-aaron committed 22 hours ago

Verified f34623bf

[Model Refactoring] Move deepseek_v4_ops to models/deepseek_v4 [3/N] (#43073)

WoosukKwon committed 22 hours ago

Verified b14be81c

Older