Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 25, 2025

• Add backward compatibility for `guided_...` API (#25615)

  Show description for 70fbdb2
  

  hmellor

  and

  DarkLight1337

  authored
  70fbdb2

  Copy full SHA for 70fbdb2

• [V0 deprecation] Remove unreachable model_config.supported_tasks (#25642)

  Show description for 7f570f1
  noooop

  authored
  7f570f1

  Copy full SHA for 7f570f1

• [Bugfix] Parse SpeculativeConfig Error (#25142)

  Show description for eaeca3c
  

  yyzxw

  and

  hmellor

  authored
  eaeca3c

  Copy full SHA for eaeca3c

• [mypy] Further improve MM type annotations (#25654)

  Show description for 12c1287
  DarkLight1337

  authored
  12c1287

  Copy full SHA for 12c1287

• [Bugfix] Fix Qwen3-VL max_num_video_tokens calculation for video profiling (#25648)

  Show description for 17b4c66
  Isotr0py

  authored
  17b4c66

  Copy full SHA for 17b4c66

• [Bugfix] Add triton.language.tensor placeholder (#25649)

  Show description for 3c2b2cc
  adobrzyn

  authored
  3c2b2cc

  Copy full SHA for 3c2b2cc

• [Misc] Fix Qwen3-VL `video_grid_thw` typing (#25646)

  Show description for 7be9ffc
  ywang96

  authored
  7be9ffc

  Copy full SHA for 7be9ffc

• [fix] Update torch version in cpu-build.txt for AArch64/ppc64le and Darwin (#25579)

  Show description for 393de22
  fadara01

  authored
  393de22

  Copy full SHA for 393de22

• Revert "[Performance] Move apply_w8a8_block_fp8_linear to an op class… (#25607)

  Show description for 1260180
  tlrmchlsmth

  authored
  1260180

  Copy full SHA for 1260180

• typo: remove duplicate `is` (#25641)

  Show description for af4ee63
  nicole-lihui

  authored
  af4ee63

  Copy full SHA for af4ee63

• Map CwmForCausalLM to llama and LlamaForCausalLM (#25611)

  Show description for bc092ea
  

  jacobkahn

  and

  ywang96

  authored
  bc092ea

  Copy full SHA for bc092ea

• [Misc] Simplify PoolerOutput and move to `v1/outputs` (#25629)

  Show description for 755ed7b
  DarkLight1337

  authored
  755ed7b

  Copy full SHA for 755ed7b

• [Bugfix] fix apply_temperature to avoid nan in probs (#24734)

  Show description for a676e66
  courage17340

  authored
  a676e66

  Copy full SHA for a676e66

• optimize: eliminate duplicate split_enc_dec_inputs calls (#25573)

  Show description for c85be1f
  nicole-lihui

  authored
  c85be1f

  Copy full SHA for c85be1f

• [Model] Add LongCat-Flash (#23991)

  Show description for 845adb3
  

  OftenDream

  and

  yangxurui

  authored
  845adb3

  Copy full SHA for 845adb3

• Enable Fbgemm NVFP4 on Dense models (#25609)

  Show description for 90b139c
  samanamp

  authored
  90b139c

  Copy full SHA for 90b139c

• [Bug] Dynamo Unsupported due to `BasevLLMParameter.torch_function` calling disabled super() (#25613)

  Show description for 4492e3a
  

  yewentao256

  and

  gemini-code-assist[bot]

  authored
  4492e3a

  Copy full SHA for 4492e3a

• [Kernel] Support DCP for Triton backend (#25132)

  Show description for 05c1948
  frank-wei

  authored
  05c1948

  Copy full SHA for 05c1948

Commits on Sep 24, 2025

• [Model] Improve DotsOCRForCausalLM (#25466)

  Show description for 52d0cb8
  jeejeelee

  authored
  52d0cb8

  Copy full SHA for 52d0cb8

• [MISC] replace c10::optional with std::optional (#25602)

  Show description for 5c1e496
  842974287

  authored
  5c1e496

  Copy full SHA for 5c1e496

• Improve `--help` for enhanced user experience (#24903)

  Show description for e7f27ea
  hmellor

  authored
  e7f27ea

  Copy full SHA for e7f27ea

• [Refactor] Use DeepGEMM Col Major TMA Aligned Tensor (#25517)

  Show description for 1f29141
  yewentao256

  authored
  1f29141

  Copy full SHA for 1f29141

• feat: BF16 FlashInfer Fused Cutlass MOE for Hopper and Blackwell Expert Parallel (#25503)

  Show description for 6160ba4
  djmmoss

  authored
  6160ba4

  Copy full SHA for 6160ba4

• [Logging] Improve log for when DeepEP HT disables CUDA Graphs (#25531)

  Show description for fea8006
  tlrmchlsmth

  authored
  fea8006

  Copy full SHA for fea8006

• [V0 Deprecation] Remove unused classes in attention (#25541)

  Show description for e6750d0
  WoosukKwon

  authored
  e6750d0

  Copy full SHA for e6750d0

• [Docs] Enable `fail_on_warning` for the docs build in CI (#25580)

  Show description for 8c85305
  hmellor

  authored
  8c85305

  Copy full SHA for 8c85305

• Suppress benign cuBLAS warning when capturing cudagraphs with DBO (#25596)

  Show description for f84a472
  SageMoore

  authored
  f84a472

  Copy full SHA for f84a472

• Support mnnvl all2allv from Flashinfer (#21003)

  Show description for 54e42b7
  

  wenscarl

  and

  tlrmchlsmth

  authored
  54e42b7

  Copy full SHA for 54e42b7

• [Bugfix] add cache model when from object storage get model (#24764)

  Show description for 2dda3e3
  lengrongfu

  authored
  2dda3e3

  Copy full SHA for 2dda3e3

• Fixes and updates to bench_per_token_quant_fp8 (#25591)

  Show description for d83f3f7
  mgoin

  authored
  d83f3f7

  Copy full SHA for d83f3f7

• [ROCm][Build][Bugfix] Fix ROCm base docker whls installation order (#25415)

  Show description for 302eb94
  gshtras

  authored
  302eb94

  Copy full SHA for 302eb94

• [ROCm][Bugfix] Only enable +rms_norm based on aiter if not explicitly disabled (#25275)

  Show description for 487745f
  gshtras

  authored
  487745f

  Copy full SHA for 487745f

• [Misc] Improve type annotations for jsontree (#25577)

  Show description for 9313be5
  DarkLight1337

  authored
  9313be5

  Copy full SHA for 9313be5

• Move `DeviceConfig`, `ObservabilityConfig`, `SpeechToTextConfig` to their own files (#25564)

  Show description for 8938774
  hmellor

  authored
  8938774

  Copy full SHA for 8938774

• [Bugfix] Fix DeepSeekV31ToolParser to correctly parse multiple tools in non-streaming output (#25405)

  Show description for e18b714
  taohui

  authored
  e18b714

  Copy full SHA for e18b714

Pagination

Previous

Next