Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 24, 2024

• Don't try to evaluate sympy.Eq in replacement; we knew this wouldn't simplify since we are here (#136533)

  Show description for 00bc175
  

  ezyang

  authored and

  pytorchmergebot

  committed
  00bc175

  Copy full SHA for 00bc175

• Add deterministic path for CUDA `cumsum` (#136224)

  Show description for b1a02bf
  

  kurtamohler

  authored and

  pytorchmergebot

  committed
  b1a02bf

  Copy full SHA for b1a02bf

• Revert "Correctly convert Python float to float64 when passing argument as Tensor (#136413)"

  Show description for 0133fbc
  pytorchmergebot

  committed
  0133fbc

  Copy full SHA for 0133fbc

• [Inductor] Rename WrapperCodeGen to PythonWrapperCodegen (#136062)

  Show description for 95c0f74
  

  desertfire

  authored and

  pytorchmergebot

  committed
  95c0f74

  Copy full SHA for 95c0f74

• [SymmetricMemory] add support for cuStreamWriteValue32 (#136488)

  Show description for da1560c
  

  Yifu Wang

  authored and

  pytorchmergebot

  committed
  da1560c

  Copy full SHA for da1560c

• [ONNX] Unify ONNXProgram and remove the old one (#136281)

  Show description for 7c777dd
  

  justinchuby

  authored and

  pytorchmergebot

  committed
  7c777dd

  Copy full SHA for 7c777dd

• [pipelining] fix py ref cycle in stage_backward (#136507)

  Show description for dbc3356
  

  wconstab

  authored and

  pytorchmergebot

  committed
  dbc3356

  Copy full SHA for dbc3356

• Fix flexattention sympy expr printer issue (#136509)

  Show description for 7ff8e66
  

  Chillee

  authored and

  pytorchmergebot

  committed
  7ff8e66

  Copy full SHA for 7ff8e66

• [inductor][test] Check if mkl dnn bf16 is supported when using bf16 (#136290)

  Show description for 02ef5dd
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  02ef5dd

  Copy full SHA for 02ef5dd

• NJT binary pointwise broadcasting support via jagged <-> padded dense conversion (#133021)

  Show description for 888744b
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  888744b

  Copy full SHA for 888744b

• [TorchScript][tensorexpr] imbue locale for IRPrinter (#136458)

  Show description for 8ecc5f1
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  8ecc5f1

  Copy full SHA for 8ecc5f1

• [MPS] Add upsample_bicubic2d as Metal op (#136123)

  Show description for c6192f3
  

  malfet

  authored and

  pytorchmergebot

  committed
  c6192f3

  Copy full SHA for c6192f3

• [dynamo] Do not treat user defined nn module attributes static for dynamic shape infra (#136516)

  Show description for dacf0c4
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  dacf0c4

  Copy full SHA for dacf0c4

• [inductor] Enable parallel compile by default in fbcode (#136246)

  Show description for 1028ced
  

  masnesral

  authored and

  pytorchmergebot

  committed
  1028ced

  Copy full SHA for 1028ced

• Allow fx graph caching higher order operators (opt-in) (#135877)

  Show description for 9abdc62
  

  oulgen

  authored and

  pytorchmergebot

  committed
  9abdc62

  Copy full SHA for 9abdc62

• Add dtypes support in opinfo for Intel Gaudi (#132840)

  Show description for efed357
  

  ankurneog

  authored and

  pytorchmergebot

  committed
  efed357

  Copy full SHA for efed357

• Revert "Increase update_hint_regression problem size to 1000 (#136434)"

  Show description for 064093a
  pytorchmergebot

  committed
  064093a

  Copy full SHA for 064093a

• Move print_export_warning so lru_cache works (#136491)

  Show description for ebfcbe0
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  ebfcbe0

  Copy full SHA for ebfcbe0

• add tolerance changes for test_sdpa_autocast in test_nestedtensor.py (#136485)

  Show description for 44ec706
  

  Fuzzkatt

  authored and

  pytorchmergebot

  committed
  44ec706

  Copy full SHA for 44ec706

• Increase bf32 tolerances for some cdist tests in test_torch (#136315)

  Show description for eac04fe
  

  robert-hardwick

  authored and

  pytorchmergebot

  committed
  eac04fe

  Copy full SHA for eac04fe

• Disable compiled autograd for re-entrant autograd (#135795)

  Show description for 0b667c0
  

  majian4work

  authored and

  pytorchmergebot

  committed
  0b667c0

  Copy full SHA for 0b667c0

• Fix ut in internal distributed_test.py (#136251)

  Show description for 33e1080
  

  gaopengff

  authored and

  pytorchmergebot

  committed
  33e1080

  Copy full SHA for 33e1080

• Remove onnx imports in dynamo (#136334)

  Show description for 58274e4
  

  justinchuby

  authored and

  pytorchmergebot

  committed
  58274e4

  Copy full SHA for 58274e4

• Avoid changing FTZ/DAZ flags in CPP builder (#136466)

  Show description for 2a178a6
  

  isuruf

  authored and

  pytorchmergebot

  committed
  2a178a6

  Copy full SHA for 2a178a6

• tf32 off for test_noncontiguous_samples in test_ops.py (#136484)

  Show description for 6300eb1
  

  Fuzzkatt

  authored and

  pytorchmergebot

  committed
  6300eb1

  Copy full SHA for 6300eb1

• Make avoid_device_init() aware of hpu device (#136194)

  Show description for 47ebb58
  

  australopitek

  authored and

  pytorchmergebot

  committed
  47ebb58

  Copy full SHA for 47ebb58

• [Docs fix] fix syntax error in docs :torch.blackman_window (#136354)

  Show description for 54fc4f5
  

  enkilee

  authored and

  pytorchmergebot

  committed
  54fc4f5

  Copy full SHA for 54fc4f5

• Add cache logs + other minor caching cleanup (#136456)

  Show description for 9fc721d
  

  aorenste

  authored and

  pytorchmergebot

  committed
  9fc721d

  Copy full SHA for 9fc721d

• [aotd] Fix freezing API for subclasses (#136265)

  Show description for 342c031
  

  IvanKobzarev

  authored and

  pytorchmergebot

  committed
  342c031

  Copy full SHA for 342c031

• [Distributed] [11/N] Fix clang-tidy warnings in torch/csrc/distributed/ (#136439)

  Show description for f048569
  

  cyyever

  authored and

  pytorchmergebot

  committed
  f048569

  Copy full SHA for f048569

• Revert "Fix tensor.data_ptr() representation overflow (#135567)"

  Show description for 538ee7b
  pytorchmergebot

  committed
  538ee7b

  Copy full SHA for 538ee7b

• Add types to _dynamo/testing.py (#136402)

  Show description for 32727b9
  

  bobrenjc93

  authored and

  pytorchmergebot

  committed
  32727b9

  Copy full SHA for 32727b9

• [dynamo][easy] support `sys.intern` (#136081)

  Show description for 73c10a0
  

  XuehaiPan

  authored and

  pytorchmergebot

  committed
  73c10a0

  Copy full SHA for 73c10a0

• deprecated datetime.utcnow() fix and _RendezvousJoinOp module initiation bug fix (#136141)

  Show description for 1266be2
  

  AminAlam

  authored and

  pytorchmergebot

  committed
  1266be2

  Copy full SHA for 1266be2

• Add option to configure reduced precision math backend for SDPA (#135964)

  Show description for 0a35986
  

  jianyuh

  authored and

  pytorchmergebot

  committed
  0a35986

  Copy full SHA for 0a35986

Pagination

Previous

Next