vllm-project/vllm

Pull Requests Commits

refactor `call_utility_consensus`

BugenZhao committed 6 hours ago

Verified 1c6ff554

use enum for `PauseMode`

BugenZhao committed 6 hours ago

Verified efde01a2

[Rust Frontend] Add /pause, /resume, /is_paused endpoints (#44499)

sahilsGit committed 6 hours ago

Verified 3c0b4432

[BugFix] Use served model name in gemma4 audio-tower error message (#44828)

llsj14 committed 9 hours ago

Verified 469f3dcf

[XPU][CI] Add more test cases in Intel GPU CI (#43663)

zxd1997066 committed 10 hours ago

Verified 94fcdd00

[ROCm][CI] Stabilizing teardown and timeout of flaky tests to prevent rare OOMs (#44761)

AndreasKaratzas committed 10 hours ago

Verified d9ff7e4e

[ROCm][CI] Stage C mirrors (#42793)

AndreasKaratzas committed 10 hours ago

Verified 967c5c3b

[XPU][Minor] format moe kernel name and add in kernel list (#44771)

yma11 committed 10 hours ago

Verified 54c660c3

[MM][CG] Simplify ViT CUDA graph interfaces (#44484)

shen-shanshan committed 10 hours ago

Verified 8fb02744

[XPU]feat: add DeepSeek-V4 XPU attention decode path (#42953)

majian4work committed 10 hours ago

Verified eebce657

[Bugfix]: Fix assertion in MambaManager.allocate_slots() (#39562)

Holworth committed 11 hours ago

Verified 303916e9

Added extra_repr() to pooler classes to improve debuggability (#44805)

taneem-ibrahim committed 13 hours ago

Verified 56334059

[Bugfix] Fix FunASR-Nano crash during initialization (#44215)

SunskyXH committed 13 hours ago

Verified 6124a98a

[Kernel][Test] Make kernel tests for mamba dual-HW (CUDA + XPU) (#42736)

adobrzyn committed 16 hours ago

Verified 2ed0a962

[1/N][KV-Cache Layout Refactor] Refactor DSV4 KV cache config construction (#44454)

LucasWilkinson committed 1 day ago

Verified 4dcd10eb

[ROCm][Kernel] Enable permute_cols for ROCm (#44674)

charlifu committed 1 day ago

Verified 228bcc43

Modify torch dependency in xpu.txt (#43087)

BramVanroy committed 1 day ago

Verified 3d3ba460

[Dependency] Remove stale cuDNN frontend upper bound (#42599)

mmangkad committed 1 day ago

Verified 66ecfd05

[CI] Stabilize the multi-audio OpenAI server path (#44051)

AndreasKaratzas committed 1 day ago

Verified f0f6805d

[Doc] Fix multimodal torch.compile troubleshooting to not use removed VLLM_TORCH_COMPILE_LEVEL (#44378)

DaoyuanLi2816 committed 1 day ago

Verified 15652a6b

[Bugfix][Mooncake] Fix per-group block_size/block_hash and group_idx in MooncakeStoreConnector KV events (#44103)

ivanium committed 1 day ago

Verified 51ef6888

[videoloader] implement glm46v video loader (#44417)

JaredforReal committed 1 day ago

Verified 6ac69203

[Cohere] Enable Cohere Mini Code model and update Command A-plus test registry (#44707)

Terrencezzj committed 1 day ago

Verified 1505b3d8

[feature] add index share feature for DSA MTP (#44420)

JaredforReal committed 1 day ago

Verified 32f34d39

[Bugfix] Fix benchmark_moe.py after inplace mechanism removal (#44041)

qyYue1389 committed 1 day ago

Verified 9c7f7741

[XPU] add xpu branch in compressed_tensors_moe_w4a4_mxfp4 (#44540)

zufangzhu committed 1 day ago

Verified 6181e80f

[XPU][Feature] transparent sleep mode support for XPU platform (#37149)

yma11 committed 1 day ago

Verified 3bb46975

[XPU] Support cpu kv offloading and tiering offloading on XPU platform (#36423)

chaojun-zhang committed 1 day ago

Verified 81096645

[DSV4] Decouple DS V4 Sparse MLA Metadata from DS V3.2 (#44699)

WoosukKwon committed 1 day ago

Verified 2a983c79

[ROCm][MLA] Replace torch.cat in sparse-MLA forward_mqa with fused concat_mla_q (#42838)

maeehart committed 1 day ago

Verified bc5745a0

Older