Commits vllm-project/vllm

[CI] Leave GPU headroom in the offline ColBERT plugin tests

khluu committed 22 hours ago

16dd7ed4

[vLLM IR] Deprecate ir_op_priority ahead of vLLM IR removal (#63371)

WoosukKwon committed 22 hours ago

Verified a50259bb

[Docs][Mistral] Note that required and named tool choice may not produce a call (#63178)

ShyamalShah3 committed 22 hours ago

Verified a0544b13

[ROCm][Model] Muse Spark: share one RoPE cache sized by max_model_len (#62757)

okorzh-amd committed 23 hours ago

Verified b1c66e55

[Bugfix][Mamba] Avoid re-marking committed prompt-tail boundaries as cached this step (#62841)

khyun1109 committed 23 hours ago

Verified d561af5d

[Doc] MiniMax-M3: use the autotune flag the pinned GB200 nightly accepts, fix golden-AL links (#63394)

Ankur-singh committed 23 hours ago

Verified 7105440c

[Doc] MiniMax-M3: document the B300 NVFP4 InferenceX AgentX aggregated profile (#63395)

Ankur-singh committed 23 hours ago

Verified c7c11188

[Doc] MiniMax-M3: document the H200 MXFP8 InferenceX AgentX aggregated profile (#63391)

Ankur-singh committed 23 hours ago

Verified 8b21328c

[ROCm][CI] Move NFS-affected MI300 test groups to MI355 (#62934)

djramic committed 23 hours ago

Verified 224b2f54

[entrypoints/openai] add kv params to streaming api response (#40186)

snadampal committed 1 day ago

Verified e2b67231

[CI] Automatically label Reflection Beam model changes (#62540)

jcotant-inferact committed 1 day ago

Verified 76518259

[Model] Add Domyn Edge support (#62544)

albertoveneri committed 1 day ago

Verified 08aa305f

[Bugfix][Tool Parser] Preserve large integer tool arguments (#55743)

Woolgathererer committed 1 day ago

Verified 44302539

[Spec Decode] Keep drafters within their rotary cache on the V2 model runner (#63251)

njhill committed 1 day ago

Verified 37cf7125

[Doc][ROCm] Document DeepSeek-V4.1-Flash MI355X AgentX TP2 profiles (#63344)

chunfangamd committed 1 day ago

Verified 74ba6384

[MRV2] Remove stale experimental README (#63370)

WoosukKwon committed 1 day ago

Verified 89faadf7

[Bugfix][CPU] Match GPU conv-state layout in GDN spec decode (#63339)

JaredforReal committed 1 day ago

Verified f3871c3c

[Perf][Kernel] Size cooperative_topk's single pass by row length, not stride (#62171)

haosdent committed 1 day ago

Verified 404c2b51

[CI][ROCm] Raise the MI355 MiniMax M3 MonoKernel timeout to 45 minutes (#63269)

AndreasKaratzas committed 1 day ago

Verified c0a41721

[Recipe] MiniMax-M3: add B200 NVFP4 AgentX aggregated profile (#63163)

hshrivastava-droid committed 1 day ago

Verified ecd1e125

[Bugfix] Skip prompt_embeds passthrough items in set_active_mm_loras (#63190)

Nan2018 committed 1 day ago

Verified b2e178ea

[Quantization] Honor MoE activation override for Humming MXFP4 (#62751)

mgoin committed 1 day ago

Verified 97b6a67d

[ROCm] Bump UCX to v1.23.0-rc1 so NIXL uses rocm_ipc (#62827)

AndreasKaratzas committed 1 day ago

Verified be08fa53

[Rust Frontend] Refactor rust frontend for latest `llm-multimodal` v1.14 crate (#59854)

Isotr0py committed 1 day ago

Verified aaa67b75

[CI][Rust Frontend] Run more Python server tests against the Rust frontend (#63320)

BugenZhao committed 1 day ago

Verified c569393e

[XPU][CI] Run prefix-caching tests for GDN, Mamba and pooling models on XPU (#62683)

faaany committed 1 day ago

Verified 4bceda5d

[XPU][UT] Use W8A8 FP8 linear kernel instead of W8A16 on XPU in modernbert_fp8 unit test (#63281)

RyanMa29 committed 1 day ago

Verified 1839702e

[Bugfix][Rust Frontend] LoraRequest: drop tensorizer_config_dict to match the Python wire layout (#63295)

aoshen02 committed 1 day ago

Verified 7b1d0744

[Rust Frontend] Support --return-tokens-as-token-ids (#63318)

BugenZhao committed 1 day ago

Verified ccd4e403

[Rust Frontend] Export Mooncake connector transfer metrics like Python (#63317)

BugenZhao committed 1 day ago

Verified 933f7962