Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on May 25, 2023

• inductor: align cpp floordiv with python floordiv for dyanmic shape path (pytorch#102068)

  Show description for 4882cd0
  

  XiaobingSuper

  authored and

  pytorchmergebot

  committed
  4882cd0

  Copy full SHA for 4882cd0

• [fx][2/n] Add metadata to placeholders (pytorch#102195)

  Show description for a896962
  

  s4ayub

  authored and

  pytorchmergebot

  committed
  a896962

  Copy full SHA for a896962

• [c10d] add fake pg necessary collectives (pytorch#102238)

  Show description for 7b47cd0
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  7b47cd0

  Copy full SHA for 7b47cd0

• [c10d] conslidate barrier after init logic (pytorch#102237)

  Show description for 9a19262
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  9a19262

  Copy full SHA for 9a19262

• Support resize on meta storage (pytorch#101988)

  Show description for 818d92f
  

  ezyang

  authored and

  pytorchmergebot

  committed
  818d92f

  Copy full SHA for 818d92f

• [TP] Add API logging for TP high level API (pytorch#102209)

  Show description for d4380ed
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  d4380ed

  Copy full SHA for d4380ed

• do not raise when constraint locals are not in signature (pytorch#102198)

  Show description for d4f711b
  

  avikchaudhuri

  authored and

  pytorchmergebot

  committed
  d4f711b

  Copy full SHA for d4f711b

• Add meta registrations for some foreach ops (pytorch#102225)

  Show description for 69c7f71
  

  mlazos

  authored and

  pytorchmergebot

  committed
  69c7f71

  Copy full SHA for 69c7f71

• Support unary not on lists (pytorch#102210)

  Show description for 2434a20
  

  mlazos

  authored and

  pytorchmergebot

  committed
  2434a20

  Copy full SHA for 2434a20

• Full default dict support in dynamo (pytorch#102202)

  Show description for 23dbdd9
  

  mlazos

  authored and

  pytorchmergebot

  committed
  23dbdd9

  Copy full SHA for 23dbdd9

• [FSDP] Start to generalize modules to ignore for mixed precision (pytorch#102010)

  Show description for f3e42f1
  

  rohan-varma

  authored and

  pytorchmergebot

  committed
  f3e42f1

  Copy full SHA for f3e42f1

• [partitioner] fix for rng ops (pytorch#102123)

  Show description for c2093de
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  c2093de

  Copy full SHA for c2093de

• Replace require_backend with require_backend_is_available (pytorch#101891)

  Show description for 11d1cd8
  

  H-Huang

  authored and

  pytorchmergebot

  committed
  11d1cd8

  Copy full SHA for 11d1cd8

Commits on May 24, 2023

• Fix redudant kernel generations (pytorch#102104)

  Show description for 3e08988
  

  FindHao

  authored and

  pytorchmergebot

  committed
  3e08988

  Copy full SHA for 3e08988

• Add fake process group (pytorch#102180)

  Show description for c903b12
  

  ezyang

  authored and

  pytorchmergebot

  committed
  c903b12

  Copy full SHA for c903b12

• add additional stream priority for cuda streams (pytorch#101956)

  Show description for 5da497c
  

  Natalia Gimelshein

  authored and

  pytorchmergebot

  committed
  5da497c

  Copy full SHA for 5da497c

• [inductor][decomp] Add aten._unsafe_index_put for unchecked indexing (pytorch#101812)

  Show description for ce42010
  

  peterbell10

  authored and

  pytorchmergebot

  committed
  ce42010

  Copy full SHA for ce42010

• Populate all args with fake tensor value (pytorch#102129)

  Show description for dbf6912
  

  tugsbayasgalan

  authored and

  pytorchmergebot

  committed
  dbf6912

  Copy full SHA for dbf6912

• Revert "Support resize on meta storage (pytorch#101988)"

  Show description for 210fc28
  pytorchmergebot

  committed
  210fc28

  Copy full SHA for 210fc28

• [distributed] implemented find_all_descendants (pytorch#102138)

  Show description for 06f656c
  

  moonbucks

  authored and

  pytorchmergebot

  committed
  06f656c

  Copy full SHA for 06f656c

• [dynamo][higher order op] Support nn.Module calls (pytorch#102022)

  Show description for 5d6810a
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  5d6810a

  Copy full SHA for 5d6810a

• Segmentation fault in flatbuffers when parsing malformed modules (pytorch#95221)

  Show description for e6fc7d8
  

  apach301

  authored and

  pytorchmergebot

  committed
  e6fc7d8

  Copy full SHA for e6fc7d8

• Add quantization lowering for nn.PixelShuffle and nn.PixelUnshuffle (pytorch#101926)

  Show description for 424c930
  

  Riley Dulin

  authored and

  pytorchmergebot

  committed
  424c930

  Copy full SHA for 424c930

• add ignored_states to FSDP/fully_shard (pytorch#102056)

  Show description for 956bd03
  

  zhaojuanmao

  authored and

  pytorchmergebot

  committed
  956bd03

  Copy full SHA for 956bd03

• Revert "Merge type stubs for torch.nn.parallel (pytorch#101528)"

  Show description for 023bc30
  pytorchmergebot

  committed
  023bc30

  Copy full SHA for 023bc30

• [spmd] Enable data parallel to work with non 0 batch dim (pytorch#100073)

  Show description for d316a2d
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  d316a2d

  Copy full SHA for d316a2d

• [spmd] add more decomp and fix a sharding bug (pytorch#100938)

  Show description for d378837
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  d378837

  Copy full SHA for d378837

• [spmd] Improve activation handling, factory ops and batch dim reduction (pytorch#100853)

  Show description for dd1f295
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  dd1f295

  Copy full SHA for dd1f295

• [spmd] enhance batch dim analysis of data parallel (pytorch#100852)

  Show description for 4d55ea8
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  4d55ea8

  Copy full SHA for 4d55ea8

• [spmd] by default average gradients for nccl backend (pytorch#99964)

  Show description for b2eaba6
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  b2eaba6

  Copy full SHA for b2eaba6

• [spmd] add option to preserve node types (pytorch#100072)

  Show description for 942cd12
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  942cd12

  Copy full SHA for 942cd12

• Unbind Cat pattern (pytorch#101767)

  Show description for fcf812c
  

  devashishshankar

  authored and

  pytorchmergebot

  committed
  fcf812c

  Copy full SHA for fcf812c

• Merge type stubs for torch.nn.parallel (pytorch#101528)

  Show description for 6cabc10
  

  ringohoffman

  authored and

  pytorchmergebot

  committed
  6cabc10

  Copy full SHA for 6cabc10

• Replace unsqueeze transform with stack (pytorch#101766)

  Show description for fdd2839
  

  devashishshankar

  authored and

  pytorchmergebot

  committed
  fdd2839

  Copy full SHA for fdd2839

• Replace split-squeeze pattern (pytorch#101765)

  Show description for c0d0a9f
  

  devashishshankar

  authored and

  pytorchmergebot

  committed
  c0d0a9f

  Copy full SHA for c0d0a9f

Pagination

Previous

Next