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

• Support resize on meta storage (pytorch#101988)

  Show description for 818d92f
  

  ezyang

  authored and

  pytorchmergebot

  committed
  818d92f

  Copy full SHA for 818d92f

• 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

• update thresholds for various ops in functorch/test_ops.py (pytorch#102016)

  Show description for 2763b50
  

  Fuzzkatt

  authored and

  pytorchmergebot

  committed
  2763b50

  Copy full SHA for 2763b50

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

• update tf32 thresholds in nn/test_convolution.py (pytorch#102015)

  Show description for f8896b7
  

  Fuzzkatt

  authored and

  pytorchmergebot

  committed
  f8896b7

  Copy full SHA for f8896b7

• [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

• 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

• [dynamo] Add astunparse dependency (pytorch#102120)

  Show description for e6af31a
  

  desertfire

  authored and

  pytorchmergebot

  committed
  e6af31a

  Copy full SHA for e6af31a

• 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

• torch.sparse.softmax: allow negative dim (pytorch#102172)

  Show description for 2e2a746
  

  nikitaved

  authored and

  pytorchmergebot

  committed
  2e2a746

  Copy full SHA for 2e2a746

• 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

• [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] 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

• No cpp + step current (pytorch#102001)

  Show description for 2232cce
  

  clee2000

  authored and

  pytorchmergebot

  committed
  2232cce

  Copy full SHA for 2232cce

• Unbind Cat pattern (pytorch#101767)

  Show description for fcf812c
  

  devashishshankar

  authored and

  pytorchmergebot

  committed
  fcf812c

  Copy full SHA for fcf812c

• 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

• [ONNX] Introduce FX-ONNX dispatcher (pytorch#100660)

  Show description for 0eb4f07
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  0eb4f07

  Copy full SHA for 0eb4f07

• [CUDA][DLPack] Handle legacy default streams for DLPack conversion (pytorch#101318)

  Show description for 66f6e0e
  

  eqy

  authored and

  pytorchmergebot

  committed
  66f6e0e

  Copy full SHA for 66f6e0e

• [inductor] Move two cpu tests to test_cpu_repro.py (pytorch#101887)

  Show description for 5f0463a
  

  desertfire

  authored and

  pytorchmergebot

  committed
  5f0463a

  Copy full SHA for 5f0463a

• [inductor] Added smooth_l1_loss refs (pytorch#102077)

  Show description for e3d97b6
  

  vfdev-5

  authored and

  pytorchmergebot

  committed
  e3d97b6

  Copy full SHA for e3d97b6

• fix inference_mode with torch.compile (pytorch#101219)

  Show description for ddf4f7b
  

  bdhirsh

  authored and

  pytorchmergebot

  committed
  ddf4f7b

  Copy full SHA for ddf4f7b

Pagination

Previous

Next