vllm-project/vllm

Pull Requests Commits

test: touch tilelang.py for coverage comparison

khluu committed 73 days ago

f28c233a

[Feat] Add support for per GPU worker RDMA NIC selection (#42083)

rajkiranjoshi committed 73 days ago

Verified 9090368b

Deprecate `JAISLMHeadModel` (#43784)

hmellor committed 73 days ago

Verified 085ac221

[MM][CG] Avoid over-padding Qwen2.5-VL encoder cudagraph window metadata (#42796)

huanghua1994 committed 73 days ago

Verified 9006204e

[ROCm] Enable the aiter top-k/top-p sampler by default (#43331)

JohnQinAMD committed 73 days ago

Verified ed7fe831

[Attention][AMD] Standardize kv layout to blocks first for AMD (#43660)

NickLucche committed 73 days ago

Verified 5b115bb8

Allow native KV cache dtype in Triton cache update (#43330)

mikekg committed 73 days ago

Verified 53a20886

[Rust Frontend] Optimize multimodal prompt expansion (#43670)

ricky-chaoju committed 73 days ago

Verified 09902476

Add Cosmos3 Reasoner model (#43356)

MaciejBalaNV committed 73 days ago

Verified 9aa131f9

[ROCm] Bump ROCm to 7.2.3 (#43136)

micah-wil committed 73 days ago

Verified 1b5437ce

[XPU][MoE] Add WNA16 oracle backend for GPTQ sym-int4 (xpu_fused_moe) (#41426)

jasonboukheir committed 73 days ago

Verified 3207e768

[ROCm][Perf] Support N=5 in wvSplitK skinny GEMM kernels for speculative decoding (#40687)

mgehre-amd committed 73 days ago

Verified a9ec46d4

[KV Offload] Rename `SecondaryTierManager.get_finished()` to `get_finished_jobs()` (#43870)

ronensc committed 73 days ago

Verified 4bfa0f2b

[Bugfix] Exclude Ray DP from #42585's deferred port allocation (#43864)

vadiklyutiy committed 73 days ago

Verified 5d126dd1

[Perf] Add do_not_specialize to Mamba SSD chunk kernels (#43803)

Majid-Taheri committed 73 days ago

Verified c08ebebf

[Bug] Fix `tests/distributed/test_elastic_ep.py - assert False` (#43813)

yewentao256 committed 73 days ago

Verified be4062fd

[rust] fix: aggregate `is_sleeping` and `reset_prefix_cache` across DP engines (#43429)

willamhou committed 73 days ago

Verified 577d6938

[Rust Frontend] Reduce Gemma4 tool parser args scan complexity (#43850)

BugenZhao committed 73 days ago

Verified 61a1e304

[Rust Frontend] Add `hy_v3` tool parser (#43872)

BugenZhao committed 73 days ago

Verified 3a282230

[CPU] Migrate cpu_awq into awq_marlin (#43841)

bigPYJ1151 committed 73 days ago

Verified 20d69d10

[Bugfix][ROCm] Resolve MoRI connector hangs at high concurrency (#40344)

simondanielsson committed 73 days ago

Verified 552eb819

[Model Refactoring] Remove torch compile dependency in DSv4 (#43746)

WoosukKwon committed 73 days ago

Verified 9957e4d2

Add token-offset based selective offload in OffloadConnector (#39983)

ruocco committed 73 days ago

Verified 864990e8

[Perf][KDA] Fuse gate softplus, chunk-local cumsum, and RCP_LN2 scaling (#43667)

zexplorerhj committed 73 days ago

Verified f3b2a819

[Perf] Optimize moe permute by pre-allocate buffer, 9~14% kernel performance improvement (#43014)

yewentao256 committed 73 days ago

Verified 64e12186

[BUGFIX] Multimodal benchmark with MistralTokenizer (#42965)

juliendenize committed 73 days ago

Verified 02606b0b

Fix `OlmoHybridForCausalLM` not initialising (#43846)

hmellor committed 73 days ago

Verified 19af4e6d

[EC Connector] Add shutdown API to EC Connector. (#42423)

omerpaz95 committed 73 days ago

Verified 811d8051

Log dummy DP step in iteration details (#41406)

vadiklyutiy committed 73 days ago

Verified c1c4db8b

[Feature] Add structured output and effort support to Anthropic Messages API (#42396)

chaunceyjiang committed 73 days ago

Verified d692b89c

Older