Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 9, 2025

• Implicit language-model-only mode via limit-mm-per-prompt (#22299)

  Show description for 08b751b
  

  19 peopleauthored
  08b751b

  Copy full SHA for 08b751b

• [Bugfix] Fix ModernBert cuda graph capturing in v1 (#21901)

  Show description for 429e4e2
  Isotr0py

  authored
  429e4e2

  Copy full SHA for 429e4e2

• [BugFix] [P/D] Handle lookahead token count edge-case with Eagle Spec Decoding and P/D (#22317)

  Show description for 35afe1b
  

  3 peopleauthored
  35afe1b

  Copy full SHA for 35afe1b

• [XPU] upgrade torch 2.8 on for XPU (#22300)

  Show description for 81c57f6
  jikunshang

  authored
  81c57f6

  Copy full SHA for 81c57f6

Commits on Aug 8, 2025

• Drop flaky test_healthcheck_response_time (#22539)

  Show description for 311d875
  russellb

  authored
  311d875

  Copy full SHA for 311d875

• Extract `CompilationConfig` from `config.py` (#22524)

  Show description for e3edc0a
  hmellor

  authored
  e3edc0a

  Copy full SHA for e3edc0a

• [Frontend] Add unix domain socket support (#18097)

  Show description for baece8c
  yyweiss

  authored
  baece8c

  Copy full SHA for baece8c

• [Docs] fix broken links in metrics.md (#22315)

  Show description for 2fcf6b2
  

  GuyStone

  and

  hmellor

  authored
  2fcf6b2

  Copy full SHA for 2fcf6b2

• Skip Qwen 1 in CI because remote code is no longer compatible with Transformers (#22536)

  Show description for 41b9655
  hmellor

  authored
  41b9655

  Copy full SHA for 41b9655

• [Bugfix] Update FA commit hash (#22546)

  Show description for bd875d2
  tdoublep

  authored
  bd875d2

  Copy full SHA for bd875d2

• [Misc] DeepGEMM : Avoid JIT generation in the hot-path (#22215)

  Show description for f703b92
  

  varun-sundar-rabindranath

  and

  Varun Sundar Rabindranath

  authored
  f703b92

  Copy full SHA for f703b92

• [BugFix] Fix IMA FlashMLA full cuda-graph and DP + Update FlashMLA (#21691)

  Show description for cd9b9de
  

  LucasWilkinson

  and

  yewentao256

  authored
  cd9b9de

  Copy full SHA for cd9b9de

• [gpt-oss] Support tool call and implement MCP tool server (#22427)

  Show description for fe6d825
  heheda12345

  authored
  fe6d825

  Copy full SHA for fe6d825

• [Docs] Rename “Distributed inference and serving” to “Parallelism & Scaling” (#22466)

  Show description for e290594
  crypdick

  authored
  e290594

  Copy full SHA for e290594

• [gpt-oss] guard import when triton kernel is not installed (#22529)

  Show description for f756a68
  

  zyongye

  and

  WoosukKwon

  authored
  f756a68

  Copy full SHA for f756a68

• [Benchmark] Add benchmark tool for multi turn conversations (#20267)

  pliops-daniels

  authored
  f0964e2

  Copy full SHA for f0964e2

• [gpt-oss] triton kernel mxfp4 (#22421)

  Show description for e789cad
  zyongye

  authored
  e789cad

  Copy full SHA for e789cad

• Remove exception for Python 3.8 typing from linter (#22506)

  Show description for e5ebeeb
  hmellor

  authored
  e5ebeeb

  Copy full SHA for e5ebeeb

• [Docs] Improve API docs (+small tweaks) (#22459)

  Show description for 7be7f38
  hmellor

  authored
  7be7f38

  Copy full SHA for 7be7f38

• [BugFix] Don't cancel asyncio tasks directly from destructors (#22476)

  Show description for ccdae73
  njhill

  authored
  ccdae73

  Copy full SHA for ccdae73

• [Misc] fix openai version (#22485)

  Show description for 9040639
  lengrongfu

  authored
  9040639

  Copy full SHA for 9040639

• [Misc] Begin deprecation of `get_tensor_model_*_group` (#22494)

  Show description for 43c4f3d
  DarkLight1337

  authored
  43c4f3d

  Copy full SHA for 43c4f3d

• [CI/Build] Fix multimodal tests (#22491)

  Show description for 1712543
  DarkLight1337

  authored
  1712543

  Copy full SHA for 1712543

• [bench] Fix benchmark/serve.py to ignore unavailable results (#22382)

  Show description for 808a7b6
  lk-chen

  authored
  808a7b6

  Copy full SHA for 808a7b6

• [Doc] Sleep mode documentation (#22310)

  Show description for 099c046
  

  6 peopleauthored
  099c046

  Copy full SHA for 099c046

• [bugfix] Fix Llama3/4 issues caused by FlashInfer 0.2.10 (#22426)

  Show description for af473f0
  nvpohanh

  authored
  af473f0

  Copy full SHA for af473f0

• Fix pre-commit (#22487)

  Show description for 157f9c1
  DarkLight1337

  authored
  157f9c1

  Copy full SHA for 157f9c1

• Optimize MiniCPMO mask creation with vectorized implementation (#22464)

  Show description for 6f28791
  skyloevil

  authored
  6f28791

  Copy full SHA for 6f28791

• not tie_word_embeddings for glm-4.5 and glm-4.5v (#22460)

  Show description for c152e2a
  zRzRzRzRzRzRzR

  authored
  c152e2a

  Copy full SHA for c152e2a

• [Bugfix] Fix RuntimeError: Index put requires the source and destination dtypes match (#22065)

  Show description for 17eaaef
  chaunceyjiang

  authored
  17eaaef

  Copy full SHA for 17eaaef

• [Kernel] Add support for block FP8 on SM120 (NVIDIA 5090 and RTX PRO 6000) (#22131)

  Show description for 3303f13
  0xjunhao

  authored
  3303f13

  Copy full SHA for 3303f13

• Fix Flashinfer CUTLASS MOE Allgather (#21963)

  Show description for b2c8ce5
  wenscarl

  authored
  b2c8ce5

  Copy full SHA for b2c8ce5

• Support Tensorrt-LLM MoE fp4 for low-latency (#21331)

  Show description for a3b9c17
  

  wenscarl

  and

  xinli-sw

  authored
  a3b9c17

  Copy full SHA for a3b9c17

• Add ModelOpt Qwen3 nvfp4 support (#20101)

  Show description for d57dc23
  Edwardf0t1

  authored
  d57dc23

  Copy full SHA for d57dc23

• [PERF] Use pybase64 to more quickly decode prompt embeddings (#22469)

  Show description for e2c8f1e
  qthequartermasterman

  authored
  e2c8f1e

  Copy full SHA for e2c8f1e

Pagination

Previous

Next