Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Mar 20, 2023

• Fix standalone compile for op with multiple outputs (#96936)

  Show description for 16d8516
  

  Valentine233

  authored and

  pytorchmergebot

  committed
  16d8516

  Copy full SHA for 16d8516

• enable Half for cat serial kernel (#96021)

  Show description for 4a99b4f
  

  jianyuh

  authored and

  pytorchmergebot

  committed
  4a99b4f

  Copy full SHA for 4a99b4f

• Add helpful pretty pretting summaries to torch for lldb debugging (#97101)

  Show description for dba9487
  

  drisspg

  authored and

  pytorchmergebot

  committed
  dba9487

  Copy full SHA for dba9487

• [BE] Remove unnecessary dict comprehensions (#97116)

  Show description for 5471621
  

  Skylion007

  authored and

  pytorchmergebot

  committed
  5471621

  Copy full SHA for 5471621

Commits on Mar 19, 2023

• [ONNX] Set shape/type into torchscript (#96349)

  Show description for be0b415
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  be0b415

  Copy full SHA for be0b415

• Replace source check with assert (#95640)

  Show description for 722c4e5
  

  voznesenskym

  authored and

  pytorchmergebot

  committed
  722c4e5

  Copy full SHA for 722c4e5

• Revert "Update mkl_verbose return value check due to API change in mkl (#96283)"

  Show description for c8030b5
  pytorchmergebot

  committed
  c8030b5

  Copy full SHA for c8030b5

• rexnet_100 is disabled for static, does not need dynamic listing (#97100)

  Show description for e74c5e5
  

  ezyang

  authored and

  pytorchmergebot

  committed
  e74c5e5

  Copy full SHA for e74c5e5

• Revert "Fix standalone compile for op with multiple outputs (#96936)"

  Show description for 5d33f9c
  pytorchmergebot

  committed
  5d33f9c

  Copy full SHA for 5d33f9c

• Update FlashAttention to work with sm90 Gpus (#97051)

  Show description for 90537a7
  

  drisspg

  authored and

  pytorchmergebot

  committed
  90537a7

  Copy full SHA for 90537a7

• Fix standalone compile for op with multiple outputs (#96936)

  Show description for 37cde56
  

  Valentine233

  authored and

  pytorchmergebot

  committed
  37cde56

  Copy full SHA for 37cde56

Commits on Mar 18, 2023

• Update mkl_verbose return value check due to API change in mkl (#96283)

  Show description for c1214ce
  

  jingxu10

  authored and

  pytorchmergebot

  committed
  c1214ce

  Copy full SHA for c1214ce

• [aot] disable inference view tracking (#96478)

  Show description for 5ee5a16
  

  Valentine233

  authored and

  pytorchmergebot

  committed
  5ee5a16

  Copy full SHA for 5ee5a16

• [dtensor] remove unused tests and fix ci (#97064)

  Show description for 4805441
  

  wanchaol

  authored and

  pytorchmergebot

  committed
  4805441

  Copy full SHA for 4805441

• Revert "[inductor] do benchmark in sub processes for max autotuning (#96410)" (#97075)

  Show description for a5923ab
  

  huydhn

  authored and

  pytorchmergebot

  committed
  a5923ab

  Copy full SHA for a5923ab

• component-level configurable logging for dynamo, inductor, aot (#94858)

  Show description for a1c46e5
  

  mlazos

  authored and

  pytorchmergebot

  committed
  a1c46e5

  Copy full SHA for a1c46e5

• Add new parameter `materialize_grads` to torch.autograd.grad() (#97015)

  Show description for 086ce76
  

  zqigolden

  authored and

  pytorchmergebot

  committed
  086ce76

  Copy full SHA for 086ce76

• [inductor] do benchmark in sub processes for max autotuning (#96410)

  Show description for 34256bc
  

  shunting314

  authored and

  pytorchmergebot

  committed
  34256bc

  Copy full SHA for 34256bc

• Update MHA doc string (#97046)

  Show description for b132220
  

  Michael Gschwind

  authored and

  pytorchmergebot

  committed
  b132220

  Copy full SHA for b132220

• [Inductor] Eliminate redundant to_dtype node (#96650)

  Show description for 915cbf8
  

  EikanWang

  authored and

  pytorchmergebot

  committed
  915cbf8

  Copy full SHA for 915cbf8

• Use is_available instead of device_count to check for CUDA availability (#97043)

  Show description for 679dec8
  

  huydhn

  authored and

  pytorchmergebot

  committed
  679dec8

  Copy full SHA for 679dec8

• Increase the timeout value for linter calculate-docker-image (#96993)

  Show description for c62fc81
  

  huydhn

  authored and

  pytorchmergebot

  committed
  c62fc81

  Copy full SHA for c62fc81

Commits on Mar 17, 2023

• [docs] passing LogSoftmax into NLLLoss (#97001)

  Show description for b390e70
  

  jamesbraza

  authored and

  pytorchmergebot

  committed
  b390e70

  Copy full SHA for b390e70

• Remove obsolete "merge -g" flag from update_commit_hashes.py (#97033)

  Show description for 410210b
  

  kit1980

  authored and

  pytorchmergebot

  committed
  410210b

  Copy full SHA for 410210b

• Re-enable test_ops_jit on Windows (#96859) (#96931)

  Show description for db2c1ea
  

  huydhn

  authored and

  pytorchmergebot

  committed
  db2c1ea

  Copy full SHA for db2c1ea

• [dynamo][dashboard] fix triton clone step in dashboard (#96623)

  Show description for a4c706b
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  a4c706b

  Copy full SHA for a4c706b

• Make keep-going work for more than linux (#96974)

  Show description for 4a90aca
  

  janeyx99

  authored and

  pytorchmergebot

  committed
  4a90aca

  Copy full SHA for 4a90aca

• Fix CPU bitwise shifts for out-of-limit shift values (#96659)

  Show description for b59a60d
  

  bjhargrave

  authored and

  pytorchmergebot

  committed
  b59a60d

  Copy full SHA for b59a60d

• Remove unnecessary items() call in zero_grad (#97040)

  Show description for dd9ade6
  

  Skylion007

  authored and

  pytorchmergebot

  committed
  dd9ade6

  Copy full SHA for dd9ade6

• [SDPA] Remove the chunk_grad from mem-eff attention (#96880)

  Show description for 98a5cf0
  

  drisspg

  authored and

  pytorchmergebot

  committed
  98a5cf0

  Copy full SHA for 98a5cf0

• Fail fast when dynamo attempts to add unspecialized int/float as additional graph inputs (#96786)

  Show description for d4b8ed2
  

  guangy10

  authored and

  pytorchmergebot

  committed
  d4b8ed2

  Copy full SHA for d4b8ed2

• Improve size mismatch error messaging referencing mat/vet sizes (#96863)

  Show description for cea13ad
  

  kiersten-stokes

  authored and

  pytorchmergebot

  committed
  cea13ad

  Copy full SHA for cea13ad

• Bind increment_version to python (#96852)

  Show description for 985fc66
  

  albanD

  authored and

  pytorchmergebot

  committed
  985fc66

  Copy full SHA for 985fc66

• Fixed print tensor.type() issue. (#96381)

  Show description for 1983b31
  

  NmomoN

  authored and

  pytorchmergebot

  committed
  1983b31

  Copy full SHA for 1983b31

• [static-runtime] one more attempt to improve crash log readability (#96903)

  Show description for 57bb5b1
  

  tenpercent

  authored and

  pytorchmergebot

  committed
  57bb5b1

  Copy full SHA for 57bb5b1

Pagination

Previous

Next