Commits

Branch selector

 main

User selector

All users

All time

Commit history

Commits on Oct 12, 2023

• Expand NT subclass to support SAM (#109123)

  Show description for 8f90be4
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  8f90be4

  Copy full SHA for 8f90be4

• [DCP] Remove _shard_tensor() call in load_sharded_optimizer_state_dict in optimizer.py (#111096)

  Show description for e0eaa95
  

  wz337

  authored and

  pytorchmergebot

  committed
  e0eaa95

  Copy full SHA for e0eaa95

• [aot_inductor] add a test with AOTInductor + TorchScript (#111124)

  Show description for 6748a14
  

  Yang Chen

  authored and

  pytorchmergebot

  committed
  6748a14

  Copy full SHA for 6748a14

• Fix typo in mixed dtypes linear operator implementation. (#111127)

  Show description for 397deaa
  

  alexsamardzic

  authored and

  pytorchmergebot

  committed
  397deaa

  Copy full SHA for 397deaa

• Revert "[inductor] Implement Fx graph caching to improve warm compilation time. (#103453)"

  Show description for 7fbfa4e
  pytorchmergebot

  committed
  7fbfa4e

  Copy full SHA for 7fbfa4e

• Add pypi required metadata to all wheels except linux (#111042)

  Show description for f905387
  

  atalman

  authored and

  pytorchmergebot

  committed
  f905387

  Copy full SHA for f905387

• Disable cutlass_template on ROCm (#111132)

  Show description for 94c9dbf
  

  jataylo

  authored and

  pytorchmergebot

  committed
  94c9dbf

  Copy full SHA for 94c9dbf

• Reland #2 "[C10] PG observability hooks. (#108815, #110907)" (#111072)

  Show description for bb1424d
  

  wconstab

  authored and

  pytorchmergebot

  committed
  bb1424d

  Copy full SHA for bb1424d

• [BE] Enable Ruff's Flake8 PYI018 (#111101)

  Show description for dede1e9
  

  isdanni

  authored and

  pytorchmergebot

  committed
  dede1e9

  Copy full SHA for dede1e9

• Added decorator `skipRocmIfTorchInductor` and skipped failing tests (#107760)

  Show description for 53a9ac5
  

  pragupta

  authored and

  pytorchmergebot

  committed
  53a9ac5

  Copy full SHA for 53a9ac5

• [Inductor] support channel last for xpu conv in inductor layout opt path (#111018)

  Show description for 918054f
  

  guangyey

  authored and

  pytorchmergebot

  committed
  918054f

  Copy full SHA for 918054f

• fix: do not reshard parameters twice (#110948)

  Show description for 5ace912
  

  0x804d8000

  authored and

  pytorchmergebot

  committed
  5ace912

  Copy full SHA for 5ace912

• Move cuda driver exit handling from helpers to threads (#111061)

  Show description for aec0f98
  

  wconstab

  authored and

  pytorchmergebot

  committed
  aec0f98

  Copy full SHA for aec0f98

• [BE] Enable Ruff's Flake8 PYI030 (#111103)

  Show description for 2f53085
  

  isdanni

  authored and

  pytorchmergebot

  committed
  2f53085

  Copy full SHA for 2f53085

• Move at::{Refcounted,}MapAllocator to c10 (#109881)

  Show description for 68a1219
  

  peterbell10

  authored and

  pytorchmergebot

  committed
  68a1219

  Copy full SHA for 68a1219

• Revert "[export] Fix graph signature data model to list of specs. (#111017)"

  Show description for 42b89ae
  pytorchmergebot

  committed
  42b89ae

  Copy full SHA for 42b89ae

• Dynamo - config gated torch.distributed allow, exclusion for special leaf funcs (#110894)

  Show description for 395d0ea
  

  voznesenskym

  authored and

  pytorchmergebot

  committed
  395d0ea

  Copy full SHA for 395d0ea

• Use CUDA image for lintrunner (#110502)

  Show description for 4991463
  

  cyyever

  authored and

  pytorchmergebot

  committed
  4991463

  Copy full SHA for 4991463

• [Dist][ez][nit] Formatted nccl version string in startup (#111076)

  Show description for d8ad0ba
  

  rohan-varma

  authored and

  pytorchmergebot

  committed
  d8ad0ba

  Copy full SHA for d8ad0ba

• [DDP] Make _ReplicateState inherit from _State and make replicate eagerly initialized (#109647)

  Show description for b35279d
  

  fegin

  authored and

  pytorchmergebot

  committed
  b35279d

  Copy full SHA for b35279d

• Move export.constrain_as_* to torch._constrain_as_* (#110757)

  Show description for 5614023
  

  tugsbayasgalan

  authored and

  pytorchmergebot

  committed
  5614023

  Copy full SHA for 5614023

• Revert "Move export.constrain_as_* to torch._constrain_as_* (#110757)"

  Show description for 6ce3a38
  pytorchmergebot

  committed
  6ce3a38

  Copy full SHA for 6ce3a38

• [vision hash update] update the pinned vision hash (#111098)

  Show description for f0e7a91
  

  pytorchupdatebot

  authored and

  pytorchmergebot

  committed
  f0e7a91

  Copy full SHA for f0e7a91

• Allow specifiying inputs as GradientEdge in autograd APIs (#110867)

  Show description for 5e8be63
  

  albanD

  authored and

  pytorchmergebot

  committed
  5e8be63

  Copy full SHA for 5e8be63

• [export] Fix graph signature data model to list of specs. (#111017)

  Show description for 33b6950
  

  zhxchen17

  authored and

  pytorchmergebot

  committed
  33b6950

  Copy full SHA for 33b6950

• [device mesh] only check when world size > num_devices per host (#111091)

  Show description for 097defb
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  097defb

  Copy full SHA for 097defb

• Use torch._check for cat error checking (#111035)

  Show description for 9316c8b
  

  ezyang

  authored and

  pytorchmergebot

  committed
  9316c8b

  Copy full SHA for 9316c8b

• Revert 107846 and 109695 (#111099)

  Show description for 6dca81c
  

  kit1980

  authored and

  pytorchmergebot

  committed
  6dca81c

  Copy full SHA for 6dca81c

• update tensor-like to check instance for torch function impl (#111087)

  Show description for 07f0f38
  

  mlazos

  authored and

  pytorchmergebot

  committed
  07f0f38

  Copy full SHA for 07f0f38

• [TP] Validate TP mesh dim for 2D composition (#111001)

  Show description for 8e32e62
  

  wz337

  authored and

  pytorchmergebot

  committed
  8e32e62

  Copy full SHA for 8e32e62

• Bump xla_base version tag to v1.1 (#109757)

  Show description for 80ea878
  

  yeounoh

  authored and

  pytorchmergebot

  committed
  80ea878

  Copy full SHA for 80ea878

• [quant][pt2e][be] Move xnnpack quantizer tests to separate file (#111004)

  Show description for e0ddc3f
  

  jerryzh168

  authored and

  pytorchmergebot

  committed
  e0ddc3f

  Copy full SHA for e0ddc3f

• Linear Quantize (#110581)

  Show description for 8f8d8a0
  

  shubhraprakash1

  authored and

  pytorchmergebot

  committed
  8f8d8a0

  Copy full SHA for 8f8d8a0

• Add `torch.unravel_index` (#110580)

  Show description for 5292a92
  

  kurtamohler

  authored and

  pytorchmergebot

  committed
  5292a92

  Copy full SHA for 5292a92

• [aotinductor] Fail models temporarily (#111100)

  Show description for 577e3df
  

  angelayi

  authored and

  pytorchmergebot

  committed
  577e3df

  Copy full SHA for 577e3df