Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jul 25, 2023

• [inductor] Make OpenMP work in fbcode (pytorch#105777)

  Show description for 5f6c6ff
  

  bertmaher

  authored and

  pytorchmergebot

  committed
  5f6c6ff

  Copy full SHA for 5f6c6ff

• [inductor] Enable vectorization in fbcode (pytorch#105756)

  Show description for b2b1f21
  

  bertmaher

  authored and

  pytorchmergebot

  committed
  b2b1f21

  Copy full SHA for b2b1f21

• [FSDP x dynamo] simplify registry keys (pytorch#104209)

  Show description for 487a33e
  

  voznesenskym

  authored and

  pytorchmergebot

  committed
  487a33e

  Copy full SHA for 487a33e

• [ONNX] Support ONNXFakeContext with op_level_debug (pytorch#105874)

  Show description for da8de04
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  da8de04

  Copy full SHA for da8de04

• Add option to disable rewriting index hints in default global save plan (pytorch#105861)

  Show description for 1032a25
  

  jonb377

  authored and

  pytorchmergebot

  committed
  1032a25

  Copy full SHA for 1032a25

• [export] Remove eliminate_dead_code (pytorch#105875)

  Show description for 8bf253e
  

  angelayi

  authored and

  pytorchmergebot

  committed
  8bf253e

  Copy full SHA for 8bf253e

• [ROCm] reduce tolerance for triangular solve with well_conditioned set to True (pytorch#104425)

  Show description for c89aec2
  

  alugorey

  authored and

  pytorchmergebot

  committed
  c89aec2

  Copy full SHA for c89aec2

• [easy] Minor torch.load docs fix (pytorch#105876)

  Show description for 9c45894
  

  mikaylagawarecki

  authored and

  pytorchmergebot

  committed
  9c45894

  Copy full SHA for 9c45894

• add basic cuda support for float8 dtypes (pytorch#105807)

  Show description for 8b34fa5
  

  vkuzo

  authored and

  pytorchmergebot

  committed
  8b34fa5

  Copy full SHA for 8b34fa5

• [PyTorch][ET] Collect Process Groups Mapping Info (pytorch#104373)

  Show description for 3a01c05
  

  louisfeng

  authored and

  pytorchmergebot

  committed
  3a01c05

  Copy full SHA for 3a01c05

• [ONNX] Export module as function (pytorch#105618)

  Show description for 00ee38c
  

  BowenBao

  authored and

  pytorchmergebot

  committed
  00ee38c

  Copy full SHA for 00ee38c

• [ONNX] Improve shape inference for Slice (pytorch#105755)

  Show description for ec33733
  

  ilyasher

  authored and

  pytorchmergebot

  committed
  ec33733

  Copy full SHA for ec33733

• Support dynamic shapes in TritonTemplates (pytorch#105295)

  Show description for 98956c5
  

  ipiszy

  authored and

  pytorchmergebot

  committed
  98956c5

  Copy full SHA for 98956c5

• [Compiled Autograd] Move to torch::dynamo::autograd namespace (pytorch#105854)

  Show description for 26e3b40
  

  jansel

  authored and

  pytorchmergebot

  committed
  26e3b40

  Copy full SHA for 26e3b40

• Provide a refined upper bound for nonzero when original numel is static (pytorch#105843)

  Show description for 5403c77
  

  ezyang

  authored and

  pytorchmergebot

  committed
  5403c77

  Copy full SHA for 5403c77

• Slightly improve AOTAutograd logging with ViewAndMutationMeta (pytorch#105702)

  Show description for cc13734
  

  ezyang

  authored and

  pytorchmergebot

  committed
  cc13734

  Copy full SHA for cc13734

• Add meta registration for foreach_maximum_.List (pytorch#105864)

  Show description for 5fec1f9
  

  janeyx99

  authored and

  pytorchmergebot

  committed
  5fec1f9

  Copy full SHA for 5fec1f9

• [FSDP][Docs] Tidy up FSDP ctor/api docs (pytorch#105847)

  Show description for 6655b65
  

  awgu

  authored and

  pytorchmergebot

  committed
  6655b65

  Copy full SHA for 6655b65

Commits on Jul 24, 2023

• [ONNX] Passes to reuse existing fake mode if possible (pytorch#105764)

  Show description for 65bce81
  

  BowenBao

  authored and

  pytorchmergebot

  committed
  65bce81

  Copy full SHA for 65bce81

• Cleanup calculate-docker-image (pytorch#105752)

  Show description for 8047ce0
  

  huydhn

  authored and

  pytorchmergebot

  committed
  8047ce0

  Copy full SHA for 8047ce0

• [inductor] triton_utils.config_of: check for divisibility by 16, even when expr is not an Integer (pytorch#105743)

  Show description for becb8dc
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  becb8dc

  Copy full SHA for becb8dc

• [functional collective] parameterize allreduce tests (pytorch#105604)

  Show description for 8b94280
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  8b94280

  Copy full SHA for 8b94280

• Meff Attn Bias (pytorch#104310)

  Show description for 5453508
  

  drisspg

  authored and

  pytorchmergebot

  committed
  5453508

  Copy full SHA for 5453508

• [exir] Add deepcopy to ExportedProgram (pytorch#105852)

  Show description for 9d62c5f
  

  JacobSzwejbka

  authored and

  pytorchmergebot

  committed
  9d62c5f

  Copy full SHA for 9d62c5f

• Compiled autograd (pytorch#103822)

  Show description for c902b84
  

  jansel

  authored and

  pytorchmergebot

  committed
  c902b84

  Copy full SHA for c902b84

• Remove unnecessary simplification in guard_lt (pytorch#105842)

  Show description for 14304af
  

  ezyang

  authored and

  pytorchmergebot

  committed
  14304af

  Copy full SHA for 14304af

• Use hipsolver for default svd case on ROCm (pytorch#103540)

  Show description for b78341d
  

  alugorey

  authored and

  pytorchmergebot

  committed
  b78341d

  Copy full SHA for b78341d

• Strengthen ConstantVariable invariants (pytorch#105796)

  Show description for bf693f2
  

  voznesenskym

  authored and

  pytorchmergebot

  committed
  bf693f2

  Copy full SHA for bf693f2

• Add a version to this signpost so I can tell if packages have taken updates (pytorch#105735)

  Show description for d2ee3d0
  

  ezyang

  authored and

  pytorchmergebot

  committed
  d2ee3d0

  Copy full SHA for d2ee3d0

• Implement NAdamW optimizer (pytorch#103881)

  Show description for b070865
  

  janEbert

  authored and

  pytorchmergebot

  committed
  b070865

  Copy full SHA for b070865

• Add SparseCsrCPU and SparseCsrCUDA dispatch to sum.dim_IntList (pytorch#99292)

  Show description for a540435
  

  yanbing-j

  authored and

  pytorchmergebot

  committed
  a540435

  Copy full SHA for a540435

• Use the newer g5.12xlarge instead of g3.16xlarge for multigpu tests (pytorch#105759)

  Show description for fb0ffee
  

  huydhn

  authored and

  pytorchmergebot

  committed
  fb0ffee

  Copy full SHA for fb0ffee

• Tweak dynamic=False behavior (pytorch#105715)

  Show description for 3045e84
  

  ezyang

  authored and

  pytorchmergebot

  committed
  3045e84

  Copy full SHA for 3045e84

• [BE] remove deprecated attributes from distributed_c10d (pytorch#105753)

  Show description for 0ab7404
  

  H-Huang

  authored and

  pytorchmergebot

  committed
  0ab7404

  Copy full SHA for 0ab7404

• use UBSAN on some tests (pytorch#103655)

  Show description for b8eb827
  

  cyyever

  authored and

  pytorchmergebot

  committed
  b8eb827

  Copy full SHA for b8eb827

Pagination

Previous

Next