Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 16, 2025

• [Benchmark] Use truncation by default for pooling benchmarks (#26992)

  Show description for 17838e5
  DarkLight1337

  authored
  17838e5

  Copy full SHA for 17838e5

• [CI/Build] Fix AMD import failures in CI (#26841)

  Show description for 44c8555
  zhewenl

  authored
  44c8555

  Copy full SHA for 44c8555

• [Hardware][CPU][PowerPC]Disable torch.compile() in toptopk sampling (#26987)

  Show description for f7d318d
  

  Akashcodes732

  and

  Akash Kaothalkar

  authored
  f7d318d

  Copy full SHA for f7d318d

• [CI/Build] Update expected beam search output for Phi3V (#26978)

  Show description for 76f0d05
  DarkLight1337

  authored
  76f0d05

  Copy full SHA for 76f0d05

• Deepseek-v3 Batch Invariant on 8xH100 (#26609)

  Show description for 7d8975d
  

  bwasti

  and

  yewentao256

  authored
  7d8975d

  Copy full SHA for 7d8975d

• [PERF] Qwen3-next MTP speedup (change bool mask indexing to index_select / index_copy to reduce d2h) (#26437)

  Show description for 785d8b6
  vadiklyutiy

  authored
  785d8b6

  Copy full SHA for 785d8b6

• [Chore] Rename `utils` submodules (#26920)

  Show description for f6cdc9a
  DarkLight1337

  authored
  f6cdc9a

  Copy full SHA for f6cdc9a

• [DOC][XPU]update feature parity with Intel GPU (#26954)

  Show description for 509cdc0
  xuechendi

  authored
  509cdc0

  Copy full SHA for 509cdc0

• [BugFix] Work around graph partition x torch.compile cache issue (#26956)

  Show description for 9b6504c
  zou3519

  authored
  9b6504c

  Copy full SHA for 9b6504c

• [bugfix] Fix SP + PP without specifying compile size (#26955)

  Show description for e19b16d
  angelayi

  authored
  e19b16d

  Copy full SHA for e19b16d

• [BUG] Allow runai_streamer_sharded in config check (#26958)

  Show description for 582f2c6
  ahao-anyscale

  authored
  582f2c6

  Copy full SHA for 582f2c6

• [CI] Enable Blackwell Llama4 MoE tests (#26731)

  Show description for f8a0acb
  mgoin

  authored
  f8a0acb

  Copy full SHA for f8a0acb

• [ROCm][FEAT] Fuse DeepSeek shared experts into AITER fused_moe ops (#24097)

  Show description for 1317034
  

  3 peopleauthored
  1317034

  Copy full SHA for 1317034

• [Bugfix] reasoning_parser parameter handling in run_batch.py (#26225)

  Show description for 0ecc553
  

  inc-jeong

  and

  USER

  authored
  0ecc553

  Copy full SHA for 0ecc553

• [Qwen3-Next] Add tuned MoE config for Qwen3-Next FP8 on H100 tp2 (#26887)

  Show description for f96bc36
  felixzhu555

  authored
  f96bc36

  Copy full SHA for f96bc36

Commits on Oct 15, 2025

• [ci] Adjusting AMD test composition 2025-10-14 (#26852)

  Show description for 938c43e
  Alexei-V-Ivanov-AMD

  authored
  938c43e

  Copy full SHA for 938c43e

• Move query quantization to attention layer for Flashinfer & Triton. (#26534)

  Show description for 0a9ef0c
  

  adabeyta

  and

  ProExpertProg

  authored
  0a9ef0c

  Copy full SHA for 0a9ef0c

• [Bug] Temporally Disable `VLLM_ALLREDUCE_USE_SYMM_MEM` by Default (#26925)

  Show description for e5b438a
  yewentao256

  authored
  e5b438a

  Copy full SHA for e5b438a

• support flashinfer_fp4 moe for 5090 gpu (#26669)

  Show description for 0b99f5d
  

  XiaobingSuper

  and

  mgoin

  authored
  0b99f5d

  Copy full SHA for 0b99f5d

• Vectorize RMS norm variance using vectorize_read_with_alignment (#26234)

  Show description for 1f491aa
  

  bbeckca

  and

  yewentao256

  authored
  1f491aa

  Copy full SHA for 1f491aa

• [NVIDIA] Add support for cudnn fp4 gemm via flashinfer (#26107)

  Show description for de92d91
  

  kaixih

  and

  mgoin

  authored
  de92d91

  Copy full SHA for de92d91

• [Chore] Clean up CODEOWNERS (#26923)

  Show description for a106362
  WoosukKwon

  authored
  a106362

  Copy full SHA for a106362

• [ModelOpt] Remove NVFP4 MoE K%16==0 constraint (#26891)

  Show description for d796375
  XiaobingSuper

  authored
  d796375

  Copy full SHA for d796375

• [Feature]: Use pydantic validation in observability.py config (#26637)

  Show description for 14f8456
  

  cern1710

  and

  hmellor

  authored
  14f8456

  Copy full SHA for 14f8456

• Olmo 3 tool parser and tests (#26143)

  Show description for 4794c2b
  pdasigi

  authored
  4794c2b

  Copy full SHA for 4794c2b

• Lower sevarity of log when model info cache misses due to exception (#26917)

  Show description for d3cbaa0
  hmellor

  authored
  d3cbaa0

  Copy full SHA for d3cbaa0

• [Chore] Separate out `vllm.utils.async_utils` (#26913)

  Show description for 828523a
  DarkLight1337

  authored
  828523a

  Copy full SHA for 828523a

• [Chore] Separate out `vllm.utils.func` (#26904)

  Show description for 136a17f
  DarkLight1337

  authored
  136a17f

  Copy full SHA for 136a17f

• [BugFix] Patch inductor memory plan logic (#26878)

  Show description for f574383
  

  BoyuanFeng

  and

  hmellor

  authored
  f574383

  Copy full SHA for f574383

• chore: remove unused marker (#26890)

  Show description for 5d59868
  max-wittig

  authored
  5d59868

  Copy full SHA for 5d59868

• [Misc] rename torch_dtype to dtype (#26695)

  Show description for 8f4b313
  wangxiyuan

  authored
  8f4b313

  Copy full SHA for 8f4b313

• [Misc] Remove `isort` and `yapf` ignores (#26888)

  Show description for f93e348
  DarkLight1337

  authored
  f93e348

  Copy full SHA for f93e348

• [Model][2/N] Improve all pooling task | Support multi-vector retrieval (#25370)

  Show description for f54f851
  noooop

  authored
  f54f851

  Copy full SHA for f54f851

• [Lora]Load tuned multi-lora kernel configs from json files (#26319)

  Show description for d4d1a60
  

  li2haipeng

  and

  jeejeelee

  authored
  d4d1a60

  Copy full SHA for d4d1a60

• [Platform] allow platform to init dp group (#22243)

  Show description for db1764e
  wangxiyuan

  authored
  db1764e

  Copy full SHA for db1764e

Pagination

Previous

Next