Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit History

Commits on Feb 11, 2026

• Reapply [Attention][FA3] Update FA3 to include new swizzle optimization (#34043)

  Show description for c7914d3
  LucasWilkinson

  authored
  c7914d3

  Copy full SHA for c7914d3

Commits on Feb 8, 2026

• [Release 2.10] Update to Torch 2.10 - final release (#30525)

  atalman

  authored
  f97ca67

  Copy full SHA for f97ca67

Commits on Feb 5, 2026

• Revert "[Attention][FA3] Update FA3 to include new swizzle optimization" (#33841)

  ProExpertProg

  authored
  e3bf79f

  Copy full SHA for e3bf79f

Commits on Feb 4, 2026

• [CPU] Split attention dispatch by head_dim alignment (#32161)

  Show description for 4dffc5e
  R3hankhan123

  authored
  4dffc5e

  Copy full SHA for 4dffc5e

Commits on Feb 3, 2026

• [Attention][FA3] Update FA3 to include new swizzle optimization (#23465)

  Show description for 2267cb1
  LucasWilkinson

  authored
  2267cb1

  Copy full SHA for 2267cb1

Commits on Jan 24, 2026

• [CPU] Improve CPU Docker build (#30953)

  Show description for 203d0bc
  

  maryamtahhan

  and

  bigPYJ1151

  authored
  203d0bc

  Copy full SHA for 203d0bc

Commits on Jan 22, 2026

• [CPU Backend] [Perf] Accelerate tensor-parallel/data-parallel inference across NUMA domains on Arm (#32792)

  Show description for 744ef30
  fadara01

  authored
  744ef30

  Copy full SHA for 744ef30

• [FlashMLA] Update FlashMLA to expose new arguments (#32810)

  Show description for 889722f
  LucasWilkinson

  authored
  889722f

  Copy full SHA for 889722f

Commits on Jan 21, 2026

• Update FlashMLA (#32491)

  Show description for b4f64e5
  LucasWilkinson

  authored
  b4f64e5

  Copy full SHA for b4f64e5

Commits on Jan 8, 2026

• [chore] Update FA commit (#30460)

  Show description for be6a81f
  LucasWilkinson

  authored
  be6a81f

  Copy full SHA for be6a81f

Commits on Jan 5, 2026

• [BugFix] Fix architecture flags to prevent issues on SM103 (#31150)

  Show description for fdcc517
  LopezCastroRoberto

  authored
  fdcc517

  Copy full SHA for fdcc517

Commits on Dec 18, 2025

• [CPU] Refactor CPU fused MOE (#30531)

  Show description for e3ab93c
  bigPYJ1151

  authored
  e3ab93c

  Copy full SHA for e3ab93c

Commits on Dec 15, 2025

• [BUILD] use sm_100f when compiling flashmla to fix support on sm103 (#30705)

  Show description for 511e81e
  Harry-Chen

  authored
  511e81e

  Copy full SHA for 511e81e

Commits on Dec 12, 2025

• [CPU][FIX] Fix build failures on Arm CPUs with torch nightly (#30481)

  Show description for f355ad5
  fadara01

  authored
  f355ad5

  Copy full SHA for f355ad5

Commits on Dec 10, 2025

• [CMake][Build]: Remove unused ACL CMake env variables (#30339)

  Show description for 1803458
  Radu2k

  authored
  1803458

  Copy full SHA for 1803458

Commits on Nov 28, 2025

• [CI/Build]: make it possible to build with a free-threaded interpreter (#29241)

  Show description for 7c1ed45
  

  rgommers

  and

  DarkLight1337

  authored
  7c1ed45

  Copy full SHA for 7c1ed45

Commits on Nov 21, 2025

• [BugFix] Fix missing symbol triggering FA2 fallback on Hopper (#29107)

  Show description for c68c7b4
  LucasWilkinson

  authored
  c68c7b4

  Copy full SHA for c68c7b4

Commits on Nov 19, 2025

• [Attention] FlashAttention ViT support, make default backend (#28763)

  Show description for 4c23690
  MatthewBonanni

  authored
  4c23690

  Copy full SHA for 4c23690

• [CPU] Refactor CPU WNA16 (#28826)

  Show description for 20852c8
  bigPYJ1151

  authored
  20852c8

  Copy full SHA for 20852c8

• [Build] Add OpenAI triton_kernels (#28788)

  Show description for 9912b8c
  varun-sundar-rabindranath

  authored
  9912b8c

  Copy full SHA for 9912b8c

Commits on Nov 14, 2025

• [Attention] Bump FA for removed method (#28429)

  Show description for 8cc40f8
  

  MatthewBonanni

  and

  DarkLight1337

  authored
  8cc40f8

  Copy full SHA for 8cc40f8

Commits on Nov 13, 2025

• [Hardware][PowerPC] Fix fp16 compilation error for Power in cpu attention backend and bump oneDNN version (#28535)

  Show description for 86d15bf
  

  Akashcodes732

  and

  Akash Kaothalkar

  authored
  86d15bf

  Copy full SHA for 86d15bf

• [build][cmake]: Bundle static ACL and torch libgomp for CPU extension builds (#28059)

  Show description for d44fbba
  Radu2k

  authored
  d44fbba

  Copy full SHA for d44fbba

Commits on Nov 12, 2025

• [CPU] Refactor CPU attention backend (#27954)

  Show description for 7f829be
  bigPYJ1151

  authored
  7f829be

  Copy full SHA for 7f829be

Commits on Nov 10, 2025

• [FA/Chore] Bump FA version for FP8 two-level accumulation (#27889)

  Show description for 9c84ca8
  

  jmkuebler

  and

  LucasWilkinson

  authored
  9c84ca8

  Copy full SHA for 9c84ca8

Commits on Nov 8, 2025

• Refactor CPU/GPU extension targets for CMake build (#28026)

  Show description for d15afc1
  ashahba

  authored
  d15afc1

  Copy full SHA for d15afc1

Commits on Oct 27, 2025

• [cpu][perf] Fix low CPU utilization with VLLM_CPU_OMP_THREADS_BIND on AArch64 (#27415)

  Show description for a663f6a
  fadara01

  authored
  a663f6a

  Copy full SHA for a663f6a

Commits on Oct 22, 2025

• [MLA] Bump FlashMLA (#27354)

  Show description for b4fda58
  MatthewBonanni

  authored
  b4fda58

  Copy full SHA for b4fda58

Commits on Oct 21, 2025

• [Bugfix] fixes the decoding metadata of dense mla's fp8 kvcache. (#27144)

  Show description for 250fb1b
  

  sighingnow

  and

  LucasWilkinson

  authored
  250fb1b

  Copy full SHA for 250fb1b

• [cpu] Dispatch un-quantized linear to oneDNN/ACL by default for AArch64 (#27183)

  Show description for 163965d
  

  fadara01

  and

  micyan01

  authored
  163965d

  Copy full SHA for 163965d

Commits on Oct 18, 2025

• [BugFix] Fix failing gemma-3-1b-it test: `test_lm_eval_accuracy_v1_engine[google/gemma-3-1b-it]` (#27111)

  Show description for 9f020f4
  LucasWilkinson

  authored
  9f020f4

  Copy full SHA for 9f020f4

Commits on Oct 15, 2025

• [CI/Build][Bugfix] fix qutlass cmake error when set QUTLASS_SRC_DIR (#26773)

  Show description for e471d7c
  

  izhuhaoran

  and

  mgoin

  authored
  e471d7c

  Copy full SHA for e471d7c

Commits on Oct 11, 2025

• fix(nix): Allow local oneDNN path to fix vLLM CPU build failure (#26401)

  Show description for 086609d
  ihb2032

  authored
  086609d

  Copy full SHA for 086609d

• [CI/Build] Fix ppc64le CPU build and tests (#22443)

  Show description for 8f8474f
  npanpaliya

  authored
  8f8474f

  Copy full SHA for 8f8474f

Commits on Oct 10, 2025

• [Transform] [Quantization] Add QuTLASS support to vLLM (#24440)

  Show description for 96ad65b
  

  3 peopleauthored
  96ad65b

  Copy full SHA for 96ad65b