Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 15, 2025

• [Benchmarks] Include image data when ShareGPT4V dataset is used. (#22955)

  Show description for 993d3d1
  huachenheli

  authored
  993d3d1

  Copy full SHA for 993d3d1

• [FIXBUG] Correctly Apply Grammar Bitmask in Mixed Batches (#22896)

  Show description for 68af77e
  JartX

  authored
  68af77e

  Copy full SHA for 68af77e

• [BugFix] Skip the Q component for QKVParallelLinear in the case of QKVCrossParallelLinear since its width is 0 (#22369)

  Show description for 6b04039
  sstamenk

  authored
  6b04039

  Copy full SHA for 6b04039

• [V0 Deprecation] Remove advance_step (#22969)

  Show description for 1c859a1
  WoosukKwon

  authored
  1c859a1

  Copy full SHA for 1c859a1

• [Core] Allow full cudagraph with separate attention routines and orthogonal to compilation, add support for FA2 and FlashInfer (#20059)

  Show description for 74f441f
  

  3 peopleauthored
  74f441f

  Copy full SHA for 74f441f

• [Frontend] Expose do_log_stats interval to env (#22905)

  Show description for a0632a3
  

  Csrayz

  and

  hmellor

  authored
  a0632a3

  Copy full SHA for a0632a3

• [CI] Remove duplicated docs build from buildkite (#22924)

  Show description for e8b40c7
  hmellor

  authored
  e8b40c7

  Copy full SHA for e8b40c7

• [Misc] Ignore ep_kernels_workspace (#22807)

  Show description for 48f4636
  jeejeelee

  authored
  48f4636

  Copy full SHA for 48f4636

• [V1] [Hybrid] Support using float32 for state in Hybrid Models (Mamba2, Mamba1, Minimax) (#22928)

  Show description for 75531a6
  

  4 peopleauthored
  75531a6

  Copy full SHA for 75531a6

• Improve multimodal hasher performance for re-used Image prompts (#22825)

  Show description for 22341b9
  p88h

  authored
  22341b9

  Copy full SHA for 22341b9

• [MM] Allow skipping memory profiling for multimodal models. (#22950)

  Show description for 49252cf
  

  Roger Wang

  and

  gemini-code-assist[bot]

  authored
  49252cf

  Copy full SHA for 49252cf

• [Bugfix] fix cuda 12.6 and 11.8 build (#22952)

  Show description for 3e6dd40
  jinzhen-lin

  authored
  3e6dd40

  Copy full SHA for 3e6dd40

• [Bugfix] Unquote file uri before reading image (#22912)

  Show description for aa300c4
  

  sayandipdutta

  and

  DarkLight1337

  authored
  aa300c4

  Copy full SHA for aa300c4

• [V1] - Split Prefill and Decode for Mamba1 models (#22653)

  Show description for fe91ce9
  

  3 peopleauthored
  fe91ce9

  Copy full SHA for fe91ce9

• [CI] Pooling models mteb test uses enforce_eager (#22878)

  Show description for 5406ebf
  noooop

  authored
  5406ebf

  Copy full SHA for 5406ebf

• [P/D]Provide bucket algorithm rate limiter for proxy_server (#22643)

  Show description for b2c0650
  

  3 peopleauthored
  b2c0650

  Copy full SHA for b2c0650

• Revert "[ROCm][AITER] Support AITER Rope ops in RotaryEmbedding Module." (#22956)

  Show description for b2f6c24
  

  tjtanaa

  and

  vllmellm

  authored
  b2f6c24

  Copy full SHA for b2f6c24

• [Mamba] - refactor: Renamed mamba_attn to mamba2_attn (#22818)

  Show description for 3d232db
  

  Josephasafg

  and

  asafgardin

  authored
  3d232db

  Copy full SHA for 3d232db

• [Feature] Full Cuda Graph Support for Cutlass MLA and 6% E2E Throughput Improvement (#22763)

  Show description for 5c3fbfe
  yewentao256

  authored
  5c3fbfe

  Copy full SHA for 5c3fbfe

• refactor: Change scaling factors calculation for flashinfer FusedMoE (#22812)

  Show description for b4cef5e
  

  amirkl94

  and

  mgoin

  authored
  b4cef5e

  Copy full SHA for b4cef5e

• [CI Perf] Prune tests in `tests/kernels/attention/` (#22936)

  Show description for 0fe8508
  mgoin

  authored
  0fe8508

  Copy full SHA for 0fe8508

• [CI Perf] Prune tests in `tests/kernels/moe/` (#22939)

  Show description for d2b0e97
  mgoin

  authored
  d2b0e97

  Copy full SHA for d2b0e97

• [CI Perf] Prune tests in `tests/kernels/quantization/` (#22942)

  Show description for 590bddb
  mgoin

  authored
  590bddb

  Copy full SHA for 590bddb

• [BugFix] Fix port lookup in internal DP LB tests (#22252)

  Show description for ae05a6d
  njhill

  authored
  ae05a6d

  Copy full SHA for ae05a6d

• [BugFix][KVConn] Fix use of `get_required_kvcache_layout` (#22734)

  Show description for 0933f9d
  njhill

  authored
  0933f9d

  Copy full SHA for 0933f9d

• Revert "[Kernel] Add cuda kernel for gpt_oss activation" (#22948)

  simon-mo

  authored
  f1f0d2f

  Copy full SHA for f1f0d2f

• [Kernel] Add cuda kernel for gpt_oss activation (#22538)

  Show description for 81f4b96
  jeejeelee

  authored
  81f4b96

  Copy full SHA for 81f4b96

Commits on Aug 14, 2025

• [Bugfix] use flash attn on sm90 (#22933)

  Show description for 39cd09d
  

  zyongye

  and

  mgoin

  authored
  39cd09d

  Copy full SHA for 39cd09d

• [BugFix] Fix initial DP request load imbalance (#22910)

  Show description for 919234f
  njhill

  authored
  919234f

  Copy full SHA for 919234f

• [Core] Return final response for aborted requests from `AsyncLLM.generate` (#22283)

  Show description for ebcce2c
  njhill

  authored
  ebcce2c

  Copy full SHA for ebcce2c

• [Quantization]: Support compressed-tensors mixed-precision model loading (#22468)

  Show description for 4121de5
  dsikka

  authored
  4121de5

  Copy full SHA for 4121de5

• [Kernel] Add nvfp4 gemm flashinfer backends (#22346)

  Show description for 279a5f3
  

  nvjullin

  and

  mgoin

  authored
  279a5f3

  Copy full SHA for 279a5f3

• [CI] Temporarily disable flaky test (#22930)

  Show description for b8ff053
  LucasWilkinson

  authored
  b8ff053

  Copy full SHA for b8ff053

• docs: update fastsafetensors usage instructions (#22891)

  Show description for 637093a
  NirLevy98

  authored
  637093a

  Copy full SHA for 637093a

• [Kernel] [Quantization] Add MXFP4 and bias support for marlin kernel (#22428)

  Show description for 33c63e9
  

  110 peopleauthored
  33c63e9

  Copy full SHA for 33c63e9

Pagination

Previous

Next