Commits

Branch selector

 ad90ab3

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 13, 2023

• Flash Attention v2 (pytorch#105602)

  Show description for ad90ab3
  

  drisspg

  authored and

  pytorchmergebot

  committed
  ad90ab3

  Copy full SHA for ad90ab3

• optests improvements based on torchvision usage on nms (pytorch#108929)

  Show description for 55f956f
  

  ezyang

  authored and

  pytorchmergebot

  committed
  55f956f

  Copy full SHA for 55f956f

• [optests] Changed failures_dict format to json; automatic update of failures_dict (pytorch#109110)

  Show description for bfa8429
  

  zou3519

  authored and

  pytorchmergebot

  committed
  bfa8429

  Copy full SHA for bfa8429

• Check index size during decomp of index_add (pytorch#108826)

  Show description for db48bc8
  

  int3

  authored and

  pytorchmergebot

  committed
  db48bc8

  Copy full SHA for db48bc8

• Enable typechecking for _inductor/virtualized.py (pytorch#108916)

  Show description for d2d36aa
  

  int3

  authored and

  pytorchmergebot

  committed
  d2d36aa

  Copy full SHA for d2d36aa

• Revert "[dynamo] preserve some FX node metadata of GraphModules (pytorch#107067)"

  Show description for c5e7588
  pytorchmergebot

  committed
  c5e7588

  Copy full SHA for c5e7588

• torch/csrc/profiler/README.md - stubs, RecordFunction, Autograd interaction (pytorch#108470)

  Show description for aee5dec
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  aee5dec

  Copy full SHA for aee5dec

• Use user directed names for variables where possible (pytorch#109092)

  Show description for de0b18f
  

  voznesenskym

  authored and

  pytorchmergebot

  committed
  de0b18f

  Copy full SHA for de0b18f

• Forward fix lint (pytorch#109177)

  Show description for 015be4c
  

  clee2000

  authored and

  pytorchmergebot

  committed
  015be4c

  Copy full SHA for 015be4c

• Update inductor ci_expected_accuracy (pytorch#109148)

  Show description for 3d8d59e
  

  angelayi

  authored and

  pytorchmergebot

  committed
  3d8d59e

  Copy full SHA for 3d8d59e

• Fix torch._numpy.random (pytorch#108944)

  Show description for 3ac2396
  

  ev-br

  authored and

  pytorchmergebot

  committed
  3ac2396

  Copy full SHA for 3ac2396

• Symintify repeat_interleave (pytorch#109133)

  Show description for 41e5d41
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  41e5d41

  Copy full SHA for 41e5d41

• Add torch.export.register_dataclass API (pytorch#109152)

  Show description for a09539f
  

  gmagogsfm

  authored and

  pytorchmergebot

  committed
  a09539f

  Copy full SHA for a09539f

• [dtensor][4/n] don't use make_fx for strategy propagation (pytorch#108262)

  Show description for 375d2ca
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  375d2ca

  Copy full SHA for 375d2ca

• [dtensor][3/n] use dedicated TensorMeta instead of the fx one (pytorch#108261)

  Show description for 09f3e08
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  09f3e08

  Copy full SHA for 09f3e08

• [dtensor][2/n] use op overload instead of function schema (pytorch#107306)

  Show description for fc1dcfb
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  fc1dcfb

  Copy full SHA for fc1dcfb

• [DCP][Test] Fix device assignment in test/distributed/checkpoint/test_file_system_checkpoint_cpu.py (pytorch#109141)

  Show description for 48e6ffb
  

  wz337

  authored and

  pytorchmergebot

  committed
  48e6ffb

  Copy full SHA for 48e6ffb

• [ONNX] Support None in fx.args as torchlib inputs (pytorch#108708)

  Show description for 91e154f
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  91e154f

  Copy full SHA for 91e154f

• [HigherOrderOp] Support SymInt as input to body function (pytorch#108967)

  Show description for a2ff345
  

  yanboliang

  authored and

  pytorchmergebot

  committed
  a2ff345

  Copy full SHA for a2ff345

• Update SingletonSymNode to allow more comparisons (pytorch#108315)

  Show description for 4667a5c
  

  soulitzer

  authored and

  pytorchmergebot

  committed
  4667a5c

  Copy full SHA for 4667a5c

• [pytorch-vulkan] add aten::randn_like & aten::normal_ (pytorch#109075)

  Show description for a46df6e
  

  tina134

  authored and

  pytorchmergebot

  committed
  a46df6e

  Copy full SHA for a46df6e

• Make mutation test work with quantized tensors (pytorch#108935)

  Show description for e5f300f
  

  ezyang

  authored and

  pytorchmergebot

  committed
  e5f300f

  Copy full SHA for e5f300f

• [submodule] Fix eltwise share buffer issue in ideep (pytorch#108038)

  Show description for 687f027
  

  yanbing-j

  authored and

  pytorchmergebot

  committed
  687f027

  Copy full SHA for 687f027

• Add torch.distributed get_rank and get_world_size to constant_fold_functions (pytorch#109029)

  Show description for e027de2
  

  ezyang

  authored and

  pytorchmergebot

  committed
  e027de2

  Copy full SHA for e027de2

• Record and replay for ShapeEnv. (pytorch#107989)

  Show description for 12e8530
  

  ysiraichi

  authored and

  pytorchmergebot

  committed
  12e8530

  Copy full SHA for 12e8530

Commits on Sep 12, 2023

• fix 'Node' object is not iterable in functorch.compile.minifier (pytorch#103011)

  Show description for e066056
  

  makslevental

  authored and

  pytorchmergebot

  committed
  e066056

  Copy full SHA for e066056

• Add memory overlap check to meta_copy_ (pytorch#108989)

  Show description for 063a626
  

  int3

  authored and

  pytorchmergebot

  committed
  063a626

  Copy full SHA for 063a626

• Fix cumprod f16 opinfo test via ref-in-float + increasing tolerances (pytorch#109128)

  Show description for f088852
  

  int3

  authored and

  pytorchmergebot

  committed
  f088852

  Copy full SHA for f088852

• Fix a bunch of opinfo tests by using reference_in_float (pytorch#109089)

  Show description for 6869b25
  

  int3

  authored and

  pytorchmergebot

  committed
  6869b25

  Copy full SHA for 6869b25

• Fix std_mean f16 opinfo test by using reference_in_float (pytorch#109081)

  Show description for baefe47
  

  int3

  authored and

  pytorchmergebot

  committed
  baefe47

  Copy full SHA for baefe47

• Reland 2: Add PyObject preservation for UntypedStorage (pytorch#109039)

  Show description for 4c5e435
  

  kurtamohler

  authored and

  pytorchmergebot

  committed
  4c5e435

  Copy full SHA for 4c5e435

• [DTensor] Remove compute_local_offset from _utils.py (pytorch#109096)

  Show description for 6dc56d3
  

  wz337

  authored and

  pytorchmergebot

  committed
  6dc56d3

  Copy full SHA for 6dc56d3

• [quant][be] Reduce warnings in tests (pytorch#108922)

  Show description for cf26e55
  

  jerryzh168

  authored and

  pytorchmergebot

  committed
  cf26e55

  Copy full SHA for cf26e55

• assign var for "not populated" str (pytorch#108844)

  Show description for 9118073
  

  redwrasse

  authored and

  pytorchmergebot

  committed
  9118073

  Copy full SHA for 9118073

• Revert "[inductor] Lower masked_scatter on CUDA (pytorch#108803)"

  Show description for 91aab16
  pytorchmergebot

  committed
  91aab16

  Copy full SHA for 91aab16

Pagination

Previous

Next