Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Jun 29, 2023

• got pytesting to work but it fails

  lengstrom

  committed
  7b41e64

  Copy full SHA for 7b41e64

• added whitespace

  lengstrom

  committed
  481b634

  Copy full SHA for 481b634

• Merge branch 'pytorch:main' into main

  lengstrom

  authored
  93d929e

  Copy full SHA for 93d929e

• added test case

  lengstrom

  committed
  7d8d3fd

  Copy full SHA for 7d8d3fd

• s390x SIMD: propagate NaN value in clamp functions (#102978)

  Show description for 732067e
  

  AlekseiNikiforovIBM

  authored and

  pytorchmergebot

  committed
  732067e

  Copy full SHA for 732067e

• Make `torch._dynamo` lazy-importable (#104368)

  Show description for fea6834
  

  malfet

  authored and

  pytorchmergebot

  committed
  fea6834

  Copy full SHA for fea6834

Commits on Jun 28, 2023

• Translation validator for dynamo guards. (#102563)

  Show description for d0a72ec
  

  ysiraichi

  authored and

  pytorchmergebot

  committed
  d0a72ec

  Copy full SHA for d0a72ec

• [RFC]: Add test for graph partition after assertion ops functionalization. (#104287)

  Show description for 3707fbf
  

  XuanQi

  authored and

  pytorchmergebot

  committed
  3707fbf

  Copy full SHA for 3707fbf

• Add gelu vulkan function (#102762)

  Show description for ede1f99
  

  shubhraprakash1

  authored and

  pytorchmergebot

  committed
  ede1f99

  Copy full SHA for ede1f99

• Replace torch.has_cuda() call with torch.backends.cuda.built() (#104338)

  Show description for f2ea27e
  

  FFFrog

  authored and

  pytorchmergebot

  committed
  f2ea27e

  Copy full SHA for f2ea27e

• Fixes ROCM_HOME detection in case no hipcc is found in path (#95634)

  Show description for e140c9c
  

  wuxxin

  authored and

  pytorchmergebot

  committed
  e140c9c

  Copy full SHA for e140c9c

• [GHF] Better check for internal diffs (#104344)

  Show description for 8464a6a
  

  malfet

  authored and

  pytorchmergebot

  committed
  8464a6a

  Copy full SHA for 8464a6a

• [dynamo] fix deep nested closure cell KeyError (#104222)

  Show description for 998c077
  

  williamwen42

  authored and

  pytorchmergebot

  committed
  998c077

  Copy full SHA for 998c077

• [inductor] convert layout of conv weight ahead of time for inference (#103642)

  Show description for 98f00f8
  

  shunting314

  authored and

  pytorchmergebot

  committed
  98f00f8

  Copy full SHA for 98f00f8

• [skip ci] Fix the deprecating link to **our office hours** (#104339)

  Show description for 044a8e3
  

  FFFrog

  authored and

  pytorchmergebot

  committed
  044a8e3

  Copy full SHA for 044a8e3

• Speed up cpp extensions re-compilation (#104280)

  Show description for b81f1d1
  

  albanD

  authored and

  pytorchmergebot

  committed
  b81f1d1

  Copy full SHA for b81f1d1

• [onnx] Convert aten::flatten with 0d input to onnx Reshape and 1d to Identity (#104089)

  Show description for 42b0bdd
  

  CYuxian

  authored and

  pytorchmergebot

  committed
  42b0bdd

  Copy full SHA for 42b0bdd

• Add DSA to IndexKernel.cu (#104054)

  Show description for aaada2c
  

  r-barnes

  authored and

  pytorchmergebot

  committed
  aaada2c

  Copy full SHA for aaada2c

• [FSDP] Check module.training for _root_cast_forward_inputs (#104223)

  Show description for c866446
  

  rohan-varma

  authored and

  pytorchmergebot

  committed
  c866446

  Copy full SHA for c866446

• Change nn.Module.__getattr__ return type to Any (#104321)

  Show description for ee19121
  

  vors

  authored and

  pytorchmergebot

  committed
  ee19121

  Copy full SHA for ee19121

• [cuDNN][cuDNN V8 API] Improve safety of `ParamsHash` keys (#104122)

  Show description for 2504af5
  

  eqy

  authored and

  pytorchmergebot

  committed
  2504af5

  Copy full SHA for 2504af5

• [Doc] Add sum reduction for CTCLoss (#100235)

  Show description for 35a8242
  

  isdanni

  authored and

  pytorchmergebot

  committed
  35a8242

  Copy full SHA for 35a8242

• [BE] Fix test_trymerge.py (#104343)

  Show description for 0a7b6dd
  

  malfet

  authored and

  pytorchmergebot

  committed
  0a7b6dd

  Copy full SHA for 0a7b6dd

• [HigherOrderOp] Fall back on all new side effects in speculate_subgraph (#104077)

  Show description for fde024b
  

  zou3519

  authored and

  pytorchmergebot

  committed
  fde024b

  Copy full SHA for fde024b

• fix specialization when you pass an unspec int into slicing on a Python list. (#104142)

  Show description for c06bb82
  

  cdzhan

  authored and

  pytorchmergebot

  committed
  c06bb82

  Copy full SHA for c06bb82

• [Easy][FSDP] Remove misleading asserts (#104274)

  Show description for 6493519
  

  Andrew Gu

  authored and

  pytorchmergebot

  committed
  6493519

  Copy full SHA for 6493519

• [FSDP] Validate `ignored_modules`, `ignored_states` (#104273)

  Show description for ba9f6e6
  

  Andrew Gu

  authored and

  pytorchmergebot

  committed
  ba9f6e6

  Copy full SHA for ba9f6e6

• [FSDP] Fix `ignored_states` doc (#104253)

  Show description for cc27e6c
  

  Andrew Gu

  authored and

  pytorchmergebot

  committed
  cc27e6c

  Copy full SHA for cc27e6c

• [FSDP] Support unfreezing params for reshard-only hook (#104186)

  Show description for 9db8ad7
  

  Andrew Gu

  authored and

  pytorchmergebot

  committed
  9db8ad7

  Copy full SHA for 9db8ad7

• [Doc] linalg.ldl_factor: render the Shape of tensor A (#99777)

  Show description for 89fcfc1
  

  isdanni

  authored and

  pytorchmergebot

  committed
  89fcfc1

  Copy full SHA for 89fcfc1

• sampled_addmm: backward performance improvements (#103544)

  Show description for 5cf3a99
  

  nikitaved

  authored and

  pytorchmergebot

  committed
  5cf3a99

  Copy full SHA for 5cf3a99

• [BE] Modernize C++ in MetalPrepackOpContext (#104312)

  Show description for 148960b
  

  malfet

  authored and

  pytorchmergebot

  committed
  148960b

  Copy full SHA for 148960b

• make funcs argument type from torch.cuda.stream as torch.Stream (#104156)

  Show description for c2095af
  

  heidongxianhua

  authored and

  pytorchmergebot

  committed
  c2095af

  Copy full SHA for c2095af

• Revert "Re-enable low memory dropout (#103330)"

  Show description for f7fdaf8
  pytorchmergebot

  committed
  f7fdaf8

  Copy full SHA for f7fdaf8

• Re-enable low memory dropout (#103330)

  Show description for 2d14395
  

  eellison

  authored and

  pytorchmergebot

  committed
  2d14395

  Copy full SHA for 2d14395

Pagination

Previous

Next