Commits vllm-project/vllm

[CI] Add explicit Buildkite step keys to 7 keyless test-area steps

khluu committed 79 days ago

66914162

fix(security): enforce VLLM_MAX_AUDIO_CLIP_FILESIZE_MB on all audio paths (#53561)

jperezdealgaba committed 79 days ago

Verified 0d7d5ed0

speculative decoding under tensor parallelism (TP>1) , workspace creation select max hidden dim of target and draft model (#52193)

khushali9 committed 79 days ago

Verified 23ab0cfd

[Refactor] Refactor batch invariance folder (#53619)

yewentao256 committed 79 days ago

Verified f59bb0bd

[Attention][Spec Decode] Support varlen trtllm-gen decode for adaptive verification (#52157)

guan404ming committed 79 days ago

Verified 6a9c69fa

[Bugfix][Multimodal] Encode text in mixed CLIP/SigLIP pooling batches (#53165)

Prudhvivuda committed 79 days ago

Verified 342b8ebd

[MISC] Cleanup deprecated parameters (#53559)

jeejeelee committed 79 days ago

Verified 4f686e18

[Core] drop duplicate VLLM_USE_DEEP_GEMM check (#48687)

andyxning committed 79 days ago

Verified 6648eb11

[Kernel][Perf] Per-architecture tuned configs for batch-invariant persistent matmul (~3x decode kernels on RTX 4090D/H20) (#53247)

LioEinaudi committed 80 days ago

Verified 7797b602

[Kimi K3 Refactor] Add `UnfinalizedMoEOutput` proto following up for #53152 (#53310)

yewentao256 committed 80 days ago

Verified 29c9af52

[Bugfix][Kimi K3] Skip absent metadata during CUDA graph profiling (#53581)

khluu committed 80 days ago

Verified 4c56e62c

[Bugfix][Spec Decode] Reapply group geometry for FlashAttention metadata (#53336)

mgoin committed 80 days ago

Verified f620499e

[Frontend] Move run_batch.py out openai folder (#53500)

noooop committed 80 days ago

Verified 9cef631f

[Bugfix][DCP] Handle sparse MLA metadata after DCP Manager refactor (#52377)

cjackal committed 80 days ago

Verified 22099afc

[MM] Cache common token sequences (#53560)

DarkLight1337 committed 80 days ago

Verified ecfa7bb3

[Bugfix] Release worker RPC payload before next dequeue (#51979)

shipiyouniao committed 80 days ago

Verified 4ca856b0

[Docs][Security] Document multimodal media UUID security implications (#53582)

jperezdealgaba committed 80 days ago

Verified 9f295fe8

[Pooling] Improve BGE-M3 sync pooling throughput by up to 3.13% (#53464)

taneem-ibrahim committed 80 days ago

Verified 79bb395e

Add Cohere ChatV2 render endpoint (#53219)

andrewbcohere committed 80 days ago

Verified 41ae917c

[Bugfix][Frontend] Fix run_batch upload retrying on success and unawaited error body (#50588)

tanchao committed 80 days ago

Verified e2399477

[Bugfix][Kernel] Handle kernel block sizes in V2 DCP slot mapping (#51031)

NickLucche committed 80 days ago

Verified 0ecc2847

Fix MNNVL Lamport mailbox publication and cleanup (#53000)

alexeldeib committed 80 days ago

Verified a047e254

[Bugfix] Reuse CUDA streams in packed weight transfer to cap reserved-memory waste (#52951)

qgallouedec committed 80 days ago

Verified 26858770

[Xeon][doc]add Xeon recipes into table (#53226)

louie-tsai committed 80 days ago

Verified cc40c367

[Model Runner V2] Reserve CUDA graph memory (#53306)

njhill committed 80 days ago

Verified a4d70bef

[Bugfix][LoRA] Add multimodal module mapping for Muse-Glimmer (#53513)

linitra24 committed 80 days ago

Verified 8c2bbe00

[LoRA] Add Qwen3-Omni multimodal LoRA support (#52786)

linitra24 committed 80 days ago

Verified 460c08bc

[Perf] Tune FlashInfer all-reduce selection on SM103 (#53318)

GirasoleY committed 80 days ago

Verified e6e1af4c

[Bugfix][XPU] Skip oneCCL warm-up all_reduce when world_size == 1 (#52389)

studioego committed 80 days ago

Verified f94666b6

[LoRA] feat: Support LoRA for DeepSeek V4 (#53361)

HollowMan6 committed 80 days ago

Verified 702e1d71