Commits

Branch selector

 main

User selector

pytorchmergebot

Datepicker

All time

Commit History

Commits on Oct 17, 2025

• Revert "Turn some const variables into constexpr in C++ code (#165401)"

  Show description for 9e94ec7
  pytorchmergebot

  committed
  9e94ec7

  Copy full SHA for 9e94ec7

• Revert "[Mem Snapshot] Add Metadata Field (#165490)"

  Show description for 11e2084
  pytorchmergebot

  committed
  11e2084

  Copy full SHA for 11e2084

• Revert "[Fix] Use sys.executable instead of hardcoded python (#165633)"

  Show description for 470e2f6
  pytorchmergebot

  committed
  470e2f6

  Copy full SHA for 470e2f6

Commits on Oct 16, 2025

• Revert "158232 Fix autocast cache incorrectly retaining no_grad state (#165068)"

  Show description for d2c82ba
  pytorchmergebot

  committed
  d2c82ba

  Copy full SHA for d2c82ba

• Revert "[inductor] print 0.0 as 0 for triton (#164291)"

  Show description for fb06e49
  pytorchmergebot

  committed
  fb06e49

  Copy full SHA for fb06e49

• Revert "[DeviceMesh] Prefer using _layout over _mesh for all sorts of things (#165554)"

  Show description for 27a98e6
  pytorchmergebot

  committed
  27a98e6

  Copy full SHA for 27a98e6

• Revert "[DeviceMesh] Introduce private constructor instead of _create_mesh_from_ranks (#165555)"

  Show description for b10f463
  pytorchmergebot

  committed
  b10f463

  Copy full SHA for b10f463

• Revert "[DeviceMesh] Simplify unflatten method (#165556)"

  Show description for 431c13c
  pytorchmergebot

  committed
  431c13c

  Copy full SHA for 431c13c

• Revert "12/n : Remove fbandroid_compiler_flags (#165558)"

  Show description for e1d71a6
  pytorchmergebot

  committed
  e1d71a6

  Copy full SHA for e1d71a6

• Revert "Add mingw to docker (#165560)"

  Show description for 69b0591
  pytorchmergebot

  committed
  69b0591

  Copy full SHA for 69b0591

Commits on Oct 15, 2025

• Revert "[Inductor][CuTeDSL] Move load_template up two directories (#165347)"

  Show description for 8c4b528
  pytorchmergebot

  committed
  8c4b528

  Copy full SHA for 8c4b528

• Revert "[inductor] Expand use of generic benchmark function (#164938)"

  Show description for 84d141e
  pytorchmergebot

  committed
  84d141e

  Copy full SHA for 84d141e

• Revert "add and fix OpInfo tests for the default partitioner (#165372)"

  Show description for b509fb9
  pytorchmergebot

  committed
  b509fb9

  Copy full SHA for b509fb9

• Revert "varlen api (#164502)"

  Show description for 3044e1a
  pytorchmergebot

  committed
  3044e1a

  Copy full SHA for 3044e1a

• Revert "[export] Handle kwargs better in aot_export_joint_with_descriptors (#165334)"

  Show description for 7778a58
  pytorchmergebot

  committed
  7778a58

  Copy full SHA for 7778a58

Commits on Oct 14, 2025

• Revert "Patch the flex_attention._get_mod_type to not use inspect.signature when computing num_positional_args (an alternative fix for flex attention graph break on create_block_mask) (#164923)"

  Show description for a2f34bd
  pytorchmergebot

  committed
  a2f34bd

  Copy full SHA for a2f34bd

• Revert "[distributed] Replace assert statements with AssertionError exceptions (#165216)"

  Show description for d2494cb
  pytorchmergebot

  committed
  d2494cb

  Copy full SHA for d2494cb

• Revert "Update windows cuda build to use 12.8 (#165345)"

  Show description for c5972eb
  pytorchmergebot

  committed
  c5972eb

  Copy full SHA for c5972eb

• Revert "use sym_numel, to allow fake tensors to work (#163831)"

  Show description for 33bfec2
  pytorchmergebot

  committed
  33bfec2

  Copy full SHA for 33bfec2

• Revert "[export] Turn on install_free_tensors flag (#164691)"

  Show description for fa3916f
  pytorchmergebot

  committed
  fa3916f

  Copy full SHA for fa3916f

• Revert "Fix double dispatch to Python for detach (#163671)"

  Show description for 267348f
  pytorchmergebot

  committed
  267348f

  Copy full SHA for 267348f

• Reapply "[export] Turn on install_free_tensors flag (#164691)" (#165353)

  Show description for 1803d40
  pytorchmergebot

  committed
  1803d40

  Copy full SHA for 1803d40

Commits on Oct 13, 2025

• Revert "[opaque_obj_v2] PyObject custom op schema type (#165004)"

  Show description for a71ca4d
  pytorchmergebot

  committed
  a71ca4d

  Copy full SHA for a71ca4d

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

  Show description for 955cd70
  pytorchmergebot

  committed
  955cd70

  Copy full SHA for 955cd70

• Revert "[dynamo][DebugMode] mask python keys in dispatch_key_set guard checks (#164992)"

  Show description for 8580112
  pytorchmergebot

  committed
  8580112

  Copy full SHA for 8580112

Commits on Oct 11, 2025

• Revert "[compile] Regional inductor compilation with fx.annotate (#164776)"

  Show description for 8d49cd5
  pytorchmergebot

  committed
  8d49cd5

  Copy full SHA for 8d49cd5

• Revert "[dynamo][annotate] Remove the need of external ctx mgr of preserve_node_meta (#165188)"

  Show description for a19123b
  pytorchmergebot

  committed
  a19123b

  Copy full SHA for a19123b

• Revert "Enable ruff rule E721 (#165162)"

  Show description for 816fb7f
  pytorchmergebot

  committed
  816fb7f

  Copy full SHA for 816fb7f

Commits on Oct 10, 2025

• Revert "Do not decompose in functionalization/proxy tensor if autograd wouldn't have decomposed (#164939)"

  Show description for 5c3fe9f
  pytorchmergebot

  committed
  5c3fe9f

  Copy full SHA for 5c3fe9f

• Revert "[dynamo][executorch] Do not trace into exeuctorch LoweredBackendModule (#165126)"

  Show description for d16627f
  pytorchmergebot

  committed
  d16627f

  Copy full SHA for d16627f

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

  Show description for b67785d
  pytorchmergebot

  committed
  b67785d

  Copy full SHA for b67785d

• Revert "Warn if AccumulateGrad stream does not match producer node stream (#165065)"

  Show description for f975bd5
  pytorchmergebot

  committed
  f975bd5

  Copy full SHA for f975bd5

• Revert "Fix truediv numerics between eager and compile (#164144)"

  Show description for abb2f71
  pytorchmergebot

  committed
  abb2f71

  Copy full SHA for abb2f71

• Revert "[AMP][Refactor] Simplify dtype support logic in autocast context manager (#163446)"

  Show description for 9420944
  pytorchmergebot

  committed
  9420944

  Copy full SHA for 9420944

• Revert "[2/N] More ruff SIM fixes (#165031)"

  Show description for b8be796
  pytorchmergebot

  committed
  b8be796

  Copy full SHA for b8be796

Pagination

Previous

Next