Commits

Branch selector

 patch-13

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 8, 2025

• Update rocm-gfx1100

  amdfaa

  authored
  b86d392

  Copy full SHA for b86d392

• remove gfx1100 from workflow

  amdfaa

  authored
  2bd470e

  Copy full SHA for 2bd470e

• Create rocm-gfx1100

  amdfaa

  authored
  a0b4652

  Copy full SHA for a0b4652

• Merge branch 'pytorch:main' into patch-13

  amdfaa

  authored
  d35a2d4

  Copy full SHA for d35a2d4

• Add split-K control to cuBLAS reduced-precision settings (#164766)

  Show description for 37c6087
  

  ngimel

  authored and

  pytorchmergebot

  committed
  37c6087

  Copy full SHA for 37c6087

• Fix refine_ranges corner case (#164075) (#164846)

  Show description for 0b85236
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  0b85236

  Copy full SHA for 0b85236

• [Max Autotune][B200] Skip carveout tests (#164435)

  Show description for 4c0fec3
  

  jananisriram

  authored and

  pytorchmergebot

  committed
  4c0fec3

  Copy full SHA for 4c0fec3

• [CMake] Remove LLVM link code (#134940)

  Show description for fdc622b
  

  cyyever

  authored and

  pytorchmergebot

  committed
  fdc622b

  Copy full SHA for fdc622b

• [ez] fix small doc error (#164915)

  Show description for 91b9484
  

  bobrenjc93

  authored and

  pytorchmergebot

  committed
  91b9484

  Copy full SHA for 91b9484

• [SymmMem] Multi-root tile reduction (#164757)

  Show description for 5c827a4
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  5c827a4

  Copy full SHA for 5c827a4

• [Benchmark] remove old timm models from benchmark (#164805)

  Show description for 8345819
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  8345819

  Copy full SHA for 8345819

• [ROCm] Improve non stride-one backwards indexing for small index sets (#164409)

  Show description for 0b01ff4
  

  doru1004

  authored and

  pytorchmergebot

  committed
  0b01ff4

  Copy full SHA for 0b01ff4

• [MPS] Update OS version in error message (#164946)

  Show description for 01f3a43
  

  malfet

  authored and

  pytorchmergebot

  committed
  01f3a43

  Copy full SHA for 01f3a43

• C++ API handle optimizer defaults (#161825)

  Show description for f332017
  

  stmcgovern

  authored and

  pytorchmergebot

  committed
  f332017

  Copy full SHA for f332017

• [PP]: Optimize memory by early releasing stage inputs' gradients (#164329)

  Show description for 0a3e4e8
  

  mingyuanw-mt

  authored and

  pytorchmergebot

  committed
  0a3e4e8

  Copy full SHA for 0a3e4e8

• Triton 3.5.x pin update to 7416ffc (#164587)

  Show description for 73adac0
  

  aakhundov

  authored and

  pytorchmergebot

  committed
  73adac0

  Copy full SHA for 73adac0

• [cuDNN][RNN] cuDNN RNN supports BFloat16 inputs since 9.13 (#164411)

  Show description for 0d39ecb
  

  eqy

  authored and

  pytorchmergebot

  committed
  0d39ecb

  Copy full SHA for 0d39ecb

• [GHF] Allow reverts from pytorch-auto-revert app (#164911)

  Show description for 90c0825
  

  malfet

  authored and

  pytorchmergebot

  committed
  90c0825

  Copy full SHA for 90c0825

• Revert "list_stored_sd_metadata API. (#160610)"

  Show description for fd4bde4
  pytorchmergebot

  committed
  fd4bde4

  Copy full SHA for fd4bde4

• Revert "Limit path search within range (#164581)"

  Show description for b5e93ff
  pytorchmergebot

  committed
  b5e93ff

  Copy full SHA for b5e93ff

• Revert "Add memory estimator (#164738)"

  Show description for f8d0d65
  pytorchmergebot

  committed
  f8d0d65

  Copy full SHA for f8d0d65

• [ROCm][CI] add gfx1150 gfx1151 to docker images for binary builds (#164854)

  Show description for f46ddb1
  

  jeffdaily

  authored and

  pytorchmergebot

  committed
  f46ddb1

  Copy full SHA for f46ddb1

• Revert "fix flex attention eager bwd: more rounding (#164317)"

  Show description for 20082d7
  pytorchmergebot

  committed
  20082d7

  Copy full SHA for 20082d7

• remove more (#164753)

  Show description for 7158aa2
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  7158aa2

  Copy full SHA for 7158aa2

• use check_size instead of check_is_size in ops.py (#164668)

  Show description for 2035f6b
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  2035f6b

  Copy full SHA for 2035f6b

• [inductor][templates] Distinguish between kernel input nodes and codegen input nodes (#163752)

  Show description for 2b58adc
  

  kundaMwiza

  authored and

  pytorchmergebot

  committed
  2b58adc

  Copy full SHA for 2b58adc

• [opaque_obj] Add make_fx tracing support (#163278)

  Show description for 322091d
  

  angelayi

  authored and

  pytorchmergebot

  committed
  322091d

  Copy full SHA for 322091d

• [opaque obj] Error for torch.library.custom_op infer_schema (#163277)

  Show description for 2bb4e68
  

  angelayi

  authored and

  pytorchmergebot

  committed
  2bb4e68

  Copy full SHA for 2bb4e68

• [opaque_obj] Add __eq__ and __deepcopy__ (#163279)

  Show description for 56ef774
  

  angelayi

  authored and

  pytorchmergebot

  committed
  56ef774

  Copy full SHA for 56ef774

• [BC-Breaking] Remove long-deprecated casting functions from native_functions.yaml (#164641)

  Show description for 64108bd
  

  cyyever

  authored and

  pytorchmergebot

  committed
  64108bd

  Copy full SHA for 64108bd

• Pyrefly suppressions 7/n (#164913)

  Show description for c855f86
  

  maggiemoss

  authored and

  pytorchmergebot

  committed
  c855f86

  Copy full SHA for c855f86

• Update round size with 1 division behavior (#162203)

  Show description for 12d2ef5
  

  morrison-turnansky

  authored and

  pytorchmergebot

  committed
  12d2ef5

  Copy full SHA for 12d2ef5

• Use codegen for the boxed interpreters (#164573)

  Show description for 65aa62d
  

  ezyang

  authored and

  pytorchmergebot

  committed
  65aa62d

  Copy full SHA for 65aa62d

• Fix #164742, all header-impl'd userfacing functions should be inline (#164871)

  Show description for 6a09f93
  

  janeyx99

  authored and

  pytorchmergebot

  committed
  6a09f93

  Copy full SHA for 6a09f93

• multimem reduce (#164517)

  Show description for 19bf67b
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  19bf67b

  Copy full SHA for 19bf67b

Pagination

Previous

Next