Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Dec 13, 2025

• [dynamo] Add ignore_fresh_unbacked_symbols for foreach ops with scalar values (#170288)

  Show description for ed1d313
  

  mlazos

  authored and

  pytorchmergebot

  committed
  ed1d313

  Copy full SHA for ed1d313

• [xpu][fix] Fix UT test_fuse_mix_order_reductions_combo_kernels (#170297)

  Show description for c267695
  

  jianyizh

  authored and

  pytorchmergebot

  committed
  c267695

  Copy full SHA for c267695

• [c10d][Sym mem] Make nccl backend full fledged with nccl 2.28.9-1 (#168129)

  Show description for c907c77
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  c907c77

  Copy full SHA for c907c77

• ci: Switch everything over to use temp envs + uv (#170316)

  Show description for b626cc1
  

  seemethere

  authored and

  pytorchmergebot

  committed
  b626cc1

  Copy full SHA for b626cc1

• ci: Move upload-test-artifacts to use uv + python (#169969)

  Show description for 9fad5c5
  

  seemethere

  authored and

  pytorchmergebot

  committed
  9fad5c5

  Copy full SHA for 9fad5c5

• [Dynamo] Fix overguarding on set/frozenset with AC (#170291)

  Show description for a52159d
  

  mlazos

  authored and

  pytorchmergebot

  committed
  a52159d

  Copy full SHA for a52159d

• [Dynamo] Fix overguarding on OrderedSet w/ AC (#169535)

  Show description for a20c119
  

  mlazos

  authored and

  pytorchmergebot

  committed
  a20c119

  Copy full SHA for a20c119

• [dynamo][DebugMode] make ModTracker a no-op in compiled regions (#170124)

  Show description for 28aa6da
  

  pianpwk

  authored and

  pytorchmergebot

  committed
  28aa6da

  Copy full SHA for 28aa6da

• Revert "Fix unsafe_chunk decomposition for empty tensors (#170053)"

  Show description for d75f366
  pytorchmergebot

  committed
  d75f366

  Copy full SHA for d75f366

• [ROCm] Enable group gemm on gfx90a (#169356)

  Show description for 5058132
  

  2 peopleauthored and

  pytorchmergebot

  committed
  5058132

  Copy full SHA for 5058132

• [18/N] Use Python 3.10 typing (#170280)

  Show description for 21705cd
  

  cyyever

  authored and

  pytorchmergebot

  committed
  21705cd

  Copy full SHA for 21705cd

• get_num_bytes small fixes (#170189)

  Show description for 0fbec1e
  

  eellison

  authored and

  pytorchmergebot

  committed
  0fbec1e

  Copy full SHA for 0fbec1e

• don't dispatch to index_put for scatter_add on integers (#170274)

  Show description for ea6685a
  

  ngimel

  authored and

  pytorchmergebot

  committed
  ea6685a

  Copy full SHA for ea6685a

• [xpu][feature] Support do bench using profiling on XPU. (#170157)

  Show description for 9ff7542
  

  etaf

  authored and

  pytorchmergebot

  committed
  9ff7542

  Copy full SHA for 9ff7542

• [inductor] Fix cudagraph skip for index_put_ with boolean indices, graph partitioning logic (#170103)

  Show description for d3008cf
  

  eellison

  authored and

  pytorchmergebot

  committed
  d3008cf

  Copy full SHA for d3008cf

• improve flex decoding perf by avoid unbalanced splitting (#167935)

  Show description for e1eb3ce
  

  shunting314

  authored and

  pytorchmergebot

  committed
  e1eb3ce

  Copy full SHA for e1eb3ce

• [invoke_subgraph] Add backend_options to nested_compile_region to be used by regional_inductor (#167599)

  Show description for cec4c4b
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  cec4c4b

  Copy full SHA for cec4c4b

• [cd] fix manylinux_2_28 tag in wheel name for aarch64 wheels (#170158)

  Show description for a09839c
  

  tinglvv

  authored and

  pytorchmergebot

  committed
  a09839c

  Copy full SHA for a09839c

• [ROCm] enable fastSpecializedAtomicAdd for gfx950 (#170330)

  Show description for 674a97c
  

  jerrymannil

  authored and

  pytorchmergebot

  committed
  674a97c

  Copy full SHA for 674a97c

• Official Docker builds forward fix. Use default python env (#170338)

  Show description for 1daa725
  

  2 peopleauthored and

  pytorchmergebot

  committed
  1daa725

  Copy full SHA for 1daa725

• Unpin patch version of nvidia-cuda-runtime for CUDA 13.0 builds (#170331)

  Show description for da03231
  

  atalman

  authored and

  pytorchmergebot

  committed
  da03231

  Copy full SHA for da03231

• Fix: torch.view_as_complex() does not work on memory layout produced by torch.contiguous() after transpose (#169780)

  Show description for 5110907
  

  redwrasse

  authored and

  pytorchmergebot

  committed
  5110907

  Copy full SHA for 5110907

• [Release 2.11] Version Bump (#170346)

  Show description for 798f420
  

  atalman

  authored and

  pytorchmergebot

  committed
  798f420

  Copy full SHA for 798f420

• [codemod][lowrisk] Remove unused exception parameter from caffe2 (#170325)

  Show description for 0213024
  

  r-barnes

  authored and

  pytorchmergebot

  committed
  0213024

  Copy full SHA for 0213024

• [CI] Update update_expected.py to skip cuda-13 results (#170348)

  Show description for 877a3a5
  

  desertfire

  authored and

  pytorchmergebot

  committed
  877a3a5

  Copy full SHA for 877a3a5

Commits on Dec 12, 2025

• [2/N] Call parent init method (#169765)

  Show description for 49f32b7
  

  cyyever

  authored and

  pytorchmergebot

  committed
  49f32b7

  Copy full SHA for 49f32b7

• Revert "[Inductor] ReLU/GELU(Addmm) fusions (#168157)"

  Show description for bdad80f
  pytorchmergebot

  committed
  bdad80f

  Copy full SHA for bdad80f

• Fix redefining of TOCH_HIP_API (#170149)

  Show description for 04f3aee
  

  cyyever

  authored and

  pytorchmergebot

  committed
  04f3aee

  Copy full SHA for 04f3aee

• Fix unsafe_chunk decomposition for empty tensors (#170053)

  Show description for d6604b8
  

  parsshar-RH

  authored and

  pytorchmergebot

  committed
  d6604b8

  Copy full SHA for d6604b8

• Remove outdated CUDA code (#170273)

  Show description for e2b0286
  

  cyyever

  authored and

  pytorchmergebot

  committed
  e2b0286

  Copy full SHA for e2b0286

• Change NamedTupleVariable implementation to subclass UserDefinedTupleVariable (#167468)

  Show description for a35ae97
  

  morrison-turnansky

  authored and

  pytorchmergebot

  committed
  a35ae97

  Copy full SHA for a35ae97

• [inductor][ez] fix the mix mm unittest (#170247)

  Show description for 23c6675
  

  shunting314

  authored and

  pytorchmergebot

  committed
  23c6675

  Copy full SHA for 23c6675

• Updating the strides to be F-continuous for nonzero_static. (#164120)

  Show description for aba14f6
  

  mansiag05

  authored and

  pytorchmergebot

  committed
  aba14f6

  Copy full SHA for aba14f6

• Revert "[dynamo] Preserve record function in aot_eager (#167787)"

  Show description for 8c5e14f
  pytorchmergebot

  committed
  8c5e14f

  Copy full SHA for 8c5e14f

• Add skip_actions flag to filter out memory snapshot events (#168183)

  Show description for dc79403
  

  jiannanWang

  authored and

  pytorchmergebot

  committed
  dc79403

  Copy full SHA for dc79403

Pagination

Previous

Next