Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Sep 17, 2025

• [Docs] vllm/benchmarks/datasets.py fix docstring param format. (#24970)

  Show description for 4a2d33e
  samzong

  authored
  4a2d33e

  Copy full SHA for 4a2d33e

• Remove old cutlass mla (#23961)

  Show description for 8f3616f
  MatthewBonanni

  authored
  8f3616f

  Copy full SHA for 8f3616f

• [Docs] improve code formatting and comments for eliminate griffe build warning. (#25010)

  Show description for 47f670b
  samzong

  authored
  47f670b

  Copy full SHA for 47f670b

• [Bugfix][Qwen3-Next] fixes the varlen issue in qwen3-next's MTP implementation. (#24957)

  Show description for dd6a910
  sighingnow

  authored
  dd6a910

  Copy full SHA for dd6a910

• [fix] lora benchmarks pass no_lora_flag_cpu (#23774)

  Show description for 1b962e2
  

  dolpm

  and

  jeejeelee

  authored
  1b962e2

  Copy full SHA for 1b962e2

• Apply fixes for CUDA 13 (#24599)

  Show description for bfe9380
  Aidyn-A

  authored
  bfe9380

  Copy full SHA for bfe9380

• [Bugfix] Fix Stream usage in CPU model runner and OneDNN kernel check (#25046)

  Show description for 9fccd04
  bigPYJ1151

  authored
  9fccd04

  Copy full SHA for 9fccd04

• Add RADIO Vision Encoder Support to vLLM (#24595)

  Show description for 252ada5
  

  danielafrimi

  and

  root

  authored
  252ada5

  Copy full SHA for 252ada5

• [Misc] Avoid use of deprecated `AutoModelForVision2Seq` (#25065)

  Show description for e120533
  DarkLight1337

  authored
  e120533

  Copy full SHA for e120533

• [BugFix] enable DOTALL to match multi-line tool_call parameters in extract_tool_call_required_streaming (#24668)

  Show description for 2b85697
  shijun-yin

  authored
  2b85697

  Copy full SHA for 2b85697

• [Frontend] Support returning all prompt logprobs (#24956)

  Show description for 544fe76
  chaunceyjiang

  authored
  544fe76

  Copy full SHA for 544fe76

• [DP] Create placement groups by ray_device_key (#25026)

  Show description for bb58dc8
  

  xinyu-intel

  and

  jikunshang

  authored
  bb58dc8

  Copy full SHA for bb58dc8

• [Docs] Fix griffe warning in base_static_graph.py (#25018)

  Show description for 0fb2551
  windsonsea

  authored
  0fb2551

  Copy full SHA for 0fb2551

• [Core] Remove tokenizer group in vLLM (#24078)

  Show description for 6c47f6b
  zhuohan123

  authored
  6c47f6b

  Copy full SHA for 6c47f6b

• [Model] Apply SharedFusedMoE to glm4_moe. (#24849)

  Show description for c15309a
  whx-sjtu

  authored
  c15309a

  Copy full SHA for c15309a

• [Model] Pass param prefix to LLMHead (#24862)

  Show description for 4a9375f
  whx-sjtu

  authored
  4a9375f

  Copy full SHA for 4a9375f

• [Core][MultiModalHasher] Hash images without converting image mode (#24969)

  Show description for 03191cd
  lgeiger

  authored
  03191cd

  Copy full SHA for 03191cd

• [EPLB] Support EPLB for Mixtral Model (#22842)

  Show description for b77bf34
  

  rouchenzi

  and

  abmfy

  authored
  b77bf34

  Copy full SHA for b77bf34

• [XPU] Fix xpu model runner call torch.cuda APIs (#25011)

  Show description for dd39baf
  jikunshang

  authored
  dd39baf

  Copy full SHA for dd39baf

• Add more documentation and improve usability of lognormal dist (benchmark_serving_multi_turn) (#23255)

  Show description for 43a62c5
  pliops-daniels

  authored
  43a62c5

  Copy full SHA for 43a62c5

• [Rocm] [quantization] Fix quark ptpc moe and add test case (#24649)

  Show description for ca2d192
  

  haoyangli0109

  and

  lihaoyang-amd

  authored
  ca2d192

  Copy full SHA for ca2d192

• [Model] Support Qwen3-VL Model Series (#24727)

  Show description for 0f7acdd
  

  4 peopleauthored
  0f7acdd

  Copy full SHA for 0f7acdd

• [V0 Deprecation] Remove MQLLMEngine (#25019)

  Show description for 5801e49
  WoosukKwon

  authored
  5801e49

  Copy full SHA for 5801e49

• [Core] Get num_encoder_tokens from scheduler config (#24989)

  Show description for 58d4c70
  russellb

  authored
  58d4c70

  Copy full SHA for 58d4c70

• [misc] fix typo in value error (#24995)

  Show description for ea3de5e
  prashantgupta24

  authored
  ea3de5e

  Copy full SHA for ea3de5e

• [UX] Remove "quantization is not fully optimized yet" log (#25012)

  Show description for 67532a1
  mgoin

  authored
  67532a1

  Copy full SHA for 67532a1

• [Docs] fix invalid doc link (#25017)

  Show description for 5672ba9
  yyzxw

  authored
  5672ba9

  Copy full SHA for 5672ba9

• [UX] Enforce valid choices for envs like VLLM_ATTENTION_BACKEND, etc (#24761)

  Show description for dd83a15
  mgoin

  authored
  dd83a15

  Copy full SHA for dd83a15

• [Benchmarks] Add MMVU video dataset support and clean up deprecated datasets (#24719)

  Show description for 5a411ef
  Isotr0py

  authored
  5a411ef

  Copy full SHA for 5a411ef

• [Core] Use `CpuGpuBuffer` for block table tensors (#24795)

  Show description for eeb135e
  njhill

  authored
  eeb135e

  Copy full SHA for eeb135e

• [Doc] Add --force-overwrite option to generate_cmake_presets.py (#24375)

  Show description for 3059b9c
  elvischenv

  authored
  3059b9c

  Copy full SHA for 3059b9c

• Removes source compilation of nixl dependency (#24874)

  Show description for 64ad551
  

  3 peopleauthored
  64ad551

  Copy full SHA for 64ad551

• [FP8] Extend per-token-group quantization support to QuantFP8 (#24342)

  Show description for cef3210
  

  3 peopleauthored
  cef3210

  Copy full SHA for cef3210

• [CI] GPT-OSS GPQA eval test for Blackwell (#24920)

  Show description for 493b10f
  

  mgoin

  and

  gemini-code-assist[bot]

  authored
  493b10f

  Copy full SHA for 493b10f

Commits on Sep 16, 2025

• [CI][Bugfix] Fix failing Blackwell test (#24993)

  Show description for d119fc8
  

  MatthewBonanni

  and

  robertgshaw2-redhat

  authored
  d119fc8

  Copy full SHA for d119fc8