vllm-project/vllm

Pull Requests Commits

Merge branch 'main' into wentao-update-agentskill

vllm-agent committed 5 hours ago

Verified 386c56bb

[Bugfix][Pooling] Honor the CrossEncoder activation saved by Sentence Transformers v6 (#62785)

cycancode committed 6 hours ago

Verified 3d3f440f

[CI][Bugfix] Read sliding_window from ManagerConfig.params in voxtral realtime test (#67476)

khluu committed 6 hours ago

Verified ae96937c

[Spec Decode] Consistent drafted, verified and accepted token counts (#64156)

tomasruizt committed 6 hours ago

Verified 157e42a7

[Model] Use FlashInfer for Gemma 4 with an NVFP4 KV cache when FA4 is unavailable (#47547)

lesj0610 committed 6 hours ago

Verified d713d1b1

[Rust Frontend] Render tool-result media in the Inkling renderer (#67398)

BugenZhao committed 6 hours ago

Verified 42a56598

[Model] Add Mistral Large 4 (#64534)

NickLucche committed 6 hours ago

Verified 25e9f7b3

[ROCm][Docs] Add DeepSeek-V4.1-Flash MI355X DP2 AgentX preset with UMBP KV offload (#67378)

shen-shanshan committed 7 hours ago

Verified 10a854ed

[Bugfix][Multimodal] Preserve values from non-native-endian NumPy embeddings (#67479)

Afloat16 committed 8 hours ago

Verified cf8fce50

[Bugfix][Benchmark] Fix prefix_repetition crash and custom_audio --output-len in bench throughput (#67288)

adtygan committed 8 hours ago

Verified 18c7d9bc

[Perf][GLM-5.3] build eager attention metadata on the host for ViT (#67194)

gcanlin committed 8 hours ago

Verified eed3ad5e

[CI][Bugfix] Migrate UMBP lazy offload tests to the ManagerConfig KV cache API (#67420)

PerryZhang01 committed 8 hours ago

Verified 65c2b39f

[Bugfix][Spec Decode] Allow min_p and logit_bias with speculative decoding (#67061)

anhtra3889 committed 10 hours ago

Verified 4dd68d13

[Perf][Kernel][Quantization] Tensor-descriptor path for int4 fused_moe_kernel_gptq_awq (XPU default, sm100 opt-in) (#52652)

afierka-intel committed 12 hours ago

Verified 191e5d5d

[Bugfix][Kernel] Bump DeepGEMM for the MegaMoE combine-readiness fix under CUDA graph replay (#67181)

gau-nernst committed 12 hours ago

Verified 3f597728

[CI] Retry the CUDA image bake when the registry rate-limits the push (#67437)

khluu committed 12 hours ago

Verified aaf0c6c9

[Rust Frontend] Keep the mock engine alive in the test_app fixture (#67409)

BugenZhao committed 12 hours ago

Verified 7d208d42

[Misc] Fix override and context manager type annotations (#67295)

hmellor committed 12 hours ago

Verified a46e2ab9

[CI] Let ci-selector decide CMakeLists.txt changes on CUDA (#65881)

khluu committed 12 hours ago

Verified 93eeae69

[Spec Decode] Keep dynamic speculative decoding for dense models with DP (#67005)

LantaoJin committed 13 hours ago

Verified 2b635e90

[ROCm][Perf][GLM-5.3-Flash] Keep the fused gated RMSNorm for the KDA o_norm (#66219)

ZhengGong-amd committed 13 hours ago

Verified 18480131

[Test][Model Loader] Leave reload headroom in the LoRA-MoE and online-quantize reload tests (#67210)

rycerzes committed 13 hours ago

Verified b98e1941

[Feature] add the option to continue serving unfinished requests after the recovery (#50265)

fangyuchu committed 14 hours ago

Verified 0ef2a95c

[Rust Frontend][gRPC] Add the Render service with RenderChat (#65994)

Dao007forever committed 14 hours ago

Verified e67fb98a

[CI][Bugfix] Pin the GLM5-next shared-expert fusion test to the stock decode (#67369)

LucasWilkinson committed 14 hours ago

Verified 329f93e1

[CPU][Perf] Split oversized final attention spans (#66855)

tianmu-li committed 14 hours ago

Verified aee2b208

[Security] Fall back to eager when encoder CUDA graph replay exceeds capture (#60112)

jperezdealgaba committed 14 hours ago

Verified ce3d4cd9

[Bugfix] Let SIGTERM during engine startup stop the API server (#67096)

abdulrahman-cohere committed 14 hours ago

Verified 5120e70e

[XPU][CI] Use platform dist backend in dist_init test fixture (#67383)

chaojun-zhang committed 14 hours ago

Verified 8d1843a9

[Bugfix] Fix M-RoPE position preparation with PCP (#67002)

pisceskkk committed 14 hours ago

Verified 74b2399b

Older