vllm-project/vllm

Pull Requests Commits

jeejeelee committed 3 days ago

42b812ed

[AMD] skip machete tests for rocm (#42326)

hissu-hyvarinen committed 3 days ago

Verified 0a62f5ee

[Bugfix][Quark] Fix W8A8 INT8 garbage outputs on Step-3.5-Flash (and other 3-key fused-MoE Quark exports) (#41892)

JoursBleu committed 3 days ago

Verified 3b1ef03b

Triton attention: add USE_TD constexpr for tensor descriptor Q/K/V load/store (#40327)

afierka-intel committed 3 days ago

Verified 3c413a54

[kv_offload] Add req_id to ReqContext for per-request tracking (#42507)

ronensc committed 3 days ago

Verified 79fd1bc7

[Bugfix][Qwen3-VL] Fix pipeline-parallel deepstack initialization (#42394)

MrZ20 committed 3 days ago

Verified cee6751e

[Bugfix] Fix scipy audio resampling ratio (#42233)

BWAAEEEK committed 3 days ago

Verified 16863072

[ROCm][CI] Skip ROCm batch invalid-input test pending torch fix (#41572)

AndreasKaratzas committed 3 days ago

Verified d628a3c5

[ROCm] Run AITER RMSNorm pad fusion before AR RMS fusion (#42411)

akii96 committed 3 days ago

Verified 74dffae6

[Bugfix][Frontend] Default max_tokens server-side on /inference/v1/generate (#42329)

hallerite committed 3 days ago

Verified 97c4317b

[CI] Use uv with Python 3.12 for PyPI wheel upload (#42470)

khluu committed 3 days ago

Verified f6e868fb

[Feat][KVConnector] Add `bind_gpu_block_pool()` to KVConnectorBase_V1 (#39654)

ivanium committed 3 days ago

Verified 13bf2421

[Bugfix] Install nvidia-cutlass-dsl[cu13] extra on CUDA 13 platforms (#42438)

ZJY0516 committed 3 days ago

Verified 140dc2ec

[Bugfix][SimpleCPUOffloadBackend] Dedup in-flight CPU offload stores across scheduler steps (#41289)

ivanium committed 3 days ago

Verified 9ce74042

Patch SlidingWindowSpec.real_page_size_bytes for nvfp4 kv (#42464)

sychen52 committed 3 days ago

Verified a8c13d28

jeejeelee committed 3 days ago

446bbb6a

jeejeelee committed 3 days ago

2e6f2b45

[MM][Perf][CG] Support ViT full CUDA graph for Qwen3.5 (#42151)

shen-shanshan committed 3 days ago

Verified 92def124

[5/n] Migrate CUTLASS MLA, hadamard, awq, allspark and DSV3 fused a gemm to torch stable ABI (continued) (#42339)

cleonard530 committed 3 days ago

Verified 85b2feca

[chore] Refactor pooling metadata token ID accessors (#42368)

taneem-ibrahim committed 3 days ago

Verified 503697c9

[Bugfix][PD] Fix multi-node TP (TP>8) (#39907)

NickLucche committed 3 days ago

Verified 71bcd02e

[Attention] Sync FA with upstream (#41052)

MatthewBonanni committed 3 days ago

Verified dcacdf9a

[MoE Refactor] Add sequence parallel tests to test_moe_layer.py (#41299)

bnellnm committed 4 days ago

Verified 18f6bf5a

[PD] Bump NIXL connector dependency to 1.x (#42364)

alec-flowers committed 4 days ago

Verified 07534b87

[Perf] Optimize MLA `compute_prefill_context` memory allocation (#42460)

yewentao256 committed 4 days ago

Verified 3d635c58

[KV Transfer] Add MooncakeStoreConnector for KV cache offloading via Mooncake distributed store (#40900)

LCAIZJ committed 4 days ago

Verified ebeb09d8

[Build] DeepGEMM: trim comments, add integration notes + TODOs (#42429)

mgoin committed 4 days ago

Verified 184577ae

[CI] Inline build artifact annotations in release pipeline (#42357)

khluu committed 4 days ago

Verified 8c4fc420

[CI] Fix `test_async_scheduling.py` flakiness (#42455)

njhill committed 4 days ago

Verified fe8b42e8

[Model Runner V2] Apply synthetic mode to probabilistic rejection sampler (#41035)

TheEpicDolphin committed 4 days ago

Verified fe5b4e0f

Older