Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 10, 2024

• Implementation of scan (#134102)

  Show description for e889252
  

  bohnstingl

  authored and

  pytorchmergebot

  committed
  e889252

  Copy full SHA for e889252

• [inductor] [cpp] fix the input contiguous check in max-autotune (#134982)

  Show description for 1469210
  

  chunyuan-w

  authored and

  pytorchmergebot

  committed
  1469210

  Copy full SHA for 1469210

Commits on Sep 9, 2024

• Fix broken E2E tests on Linux machines (#135394)

  Show description for cbc6b30
  

  maclyn

  authored and

  pytorchmergebot

  committed
  cbc6b30

  Copy full SHA for cbc6b30

• Fix bugs blocking flipping the default layout constraint for custom ops (#135391)

  Show description for 5f7d956
  

  zou3519

  authored and

  pytorchmergebot

  committed
  5f7d956

  Copy full SHA for 5f7d956

• Revert "[Inductor] Make static_input_idxs a set for faster lookup (#135314)"

  Show description for 73a6fc6
  pytorchmergebot

  committed
  73a6fc6

  Copy full SHA for 73a6fc6

• [AOTI] Fix assert_function call in cpu autotune template (#135086)

  Show description for 16c3b8f
  

  desertfire

  authored and

  pytorchmergebot

  committed
  16c3b8f

  Copy full SHA for 16c3b8f

• [AOTI] Add C shim for aten.mkldnn_rnn_layer in cpp wrapper (#134857)

  Show description for 9c6dff4
  

  desertfire

  authored and

  pytorchmergebot

  committed
  9c6dff4

  Copy full SHA for 9c6dff4

• [Inductor] Make static_input_idxs a set for faster lookup (#135314)

  Show description for 011cae9
  

  platers

  authored and

  pytorchmergebot

  committed
  011cae9

  Copy full SHA for 011cae9

• [inductor][cpp][gemm] fix perf regression xcit_large_24_p8_224 (#134686) (#135438)

  Show description for c0436c5
  

  Jiong Gong

  authored and

  pytorchmergebot

  committed
  c0436c5

  Copy full SHA for c0436c5

Commits on Sep 8, 2024

• [inductor] Cleanup analysis done at lowering time (#135412)

  Show description for 31c4e0d
  

  jansel

  authored and

  pytorchmergebot

  committed
  31c4e0d

  Copy full SHA for 31c4e0d

• [inductor] Refactor BaseSchedulerNode.__init__ (#135400)

  Show description for 53290ca
  

  jansel

  authored and

  pytorchmergebot

  committed
  53290ca

  Copy full SHA for 53290ca

• [inductor] Fast path for extract_read_writes without tracing (#135377)

  Show description for 16f5155
  

  jansel

  authored and

  pytorchmergebot

  committed
  16f5155

  Copy full SHA for 16f5155

• [inductor] Remove ReadWrites.op_counts (#135306)

  Show description for 37144be
  

  jansel

  authored and

  pytorchmergebot

  committed
  37144be

  Copy full SHA for 37144be

• [inductor] Refactor LoopBody.memory_usage (#135286)

  Show description for 3bdc54e
  

  jansel

  authored and

  pytorchmergebot

  committed
  3bdc54e

  Copy full SHA for 3bdc54e

• Change wrapped_linear_prepack and wrapped_quantized_linear_prepacked to private by adding _ as prefix (#135401)

  Show description for fd494dd
  

  hl475

  authored and

  pytorchmergebot

  committed
  fd494dd

  Copy full SHA for fd494dd

Commits on Sep 7, 2024

• [FlexAttention] Align the matmul tensorcore usage (#135168)

  Show description for 3bebc09
  

  drisspg

  authored and

  pytorchmergebot

  committed
  3bebc09

  Copy full SHA for 3bebc09

• [inductor] Catch BrokenProcessPool and print a more helpful message. (#135120)

  Show description for a2db22e
  

  masnesral

  authored and

  pytorchmergebot

  committed
  a2db22e

  Copy full SHA for a2db22e

• [inductor] Move LoopBody to its own file (#135257)

  Show description for eac5e12
  

  jansel

  authored and

  pytorchmergebot

  committed
  eac5e12

  Copy full SHA for eac5e12

• [inductor][cpp][gemm] cache blocking config for dynamic shapes (#133538)

  Show description for d7c97e7
  

  2 peopleauthored and

  pytorchmergebot

  committed
  d7c97e7

  Copy full SHA for d7c97e7

• [inductor][cpp][gemm] enable dynamic M for k-slicing (#133447)

  Show description for be9f4ff
  

  2 peopleauthored and

  pytorchmergebot

  committed
  be9f4ff

  Copy full SHA for be9f4ff

• [inductor][cpp][gemm] reduce memory alloc overhead by allocating local acc once per thread (#135277)

  Show description for 692faa9
  

  2 peopleauthored and

  pytorchmergebot

  committed
  692faa9

  Copy full SHA for 692faa9

• [Inductor] Optionally allow padding on non-GPU devices (#135280)

  Show description for e6a0221
  

  blaine-rister

  authored and

  pytorchmergebot

  committed
  e6a0221

  Copy full SHA for e6a0221

Commits on Sep 6, 2024

• [Inductor][CPP] Fix the issue of view dtype (#135301)

  Show description for 2c7e314
  

  leslie-fang-intel

  authored and

  pytorchmergebot

  committed
  2c7e314

  Copy full SHA for 2c7e314

• [inductor] Fix loop split optimization (#135303)

  Show description for ead4407
  

  jiayisunx

  authored and

  pytorchmergebot

  committed
  ead4407

  Copy full SHA for ead4407

• Run bypassed graph compile outside the except block to avoid chaining of exceptions (#135175)

  Show description for 13ba0a2
  

  oulgen

  authored and

  pytorchmergebot

  committed
  13ba0a2

  Copy full SHA for 13ba0a2

• [aoti][easy] remove breakpoint() in wrapper.py (#134807)

  Show description for 3988b34
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  3988b34

  Copy full SHA for 3988b34

• Add Inductor config for default stride behavior (#135238)

  Show description for ad29a2c
  

  zou3519

  authored and

  pytorchmergebot

  committed
  ad29a2c

  Copy full SHA for ad29a2c

• [inductor][triton] mark workspace args as mutated (#134648)

  Show description for a086882
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  a086882

  Copy full SHA for a086882

• [inductor] [cpp] improve cache blocking for is_dynamic_M (#131306)

  Show description for 13bae39
  

  2 peopleauthored and

  pytorchmergebot

  committed
  13bae39

  Copy full SHA for 13bae39

• [inductor][cpp][gemm] fix autotune runtime error from linear_binary fusion (#135275)

  Show description for 4ef6c05
  

  Jiong Gong

  authored and

  pytorchmergebot

  committed
  4ef6c05

  Copy full SHA for 4ef6c05

• [inductor] Fix gen_transposed_tile_load_store (#135307)

  Show description for 67c7924
  

  jiayisunx

  authored and

  pytorchmergebot

  committed
  67c7924

  Copy full SHA for 67c7924

• [Inductor][CPP] Select tiling factor for lower precision data types (#133830)

  Show description for 758d515
  

  CaoE

  authored and

  pytorchmergebot

  committed
  758d515

  Copy full SHA for 758d515

• [FlexAttention] Specify padding_value for boundary checked loads (#134573)

  Show description for 764ee6e
  

  drisspg

  authored and

  pytorchmergebot

  committed
  764ee6e

  Copy full SHA for 764ee6e

• [inductor] Remove LoopBody.reads,writes,other (#135256)

  Show description for 7ffb3b2
  

  jansel

  authored and

  pytorchmergebot

  committed
  7ffb3b2

  Copy full SHA for 7ffb3b2

• [inductor] Skip retracing an existing LoopBody (#135235)

  Show description for f946bf8
  

  jansel

  authored and

  pytorchmergebot

  committed
  f946bf8

  Copy full SHA for f946bf8

Pagination

Previous

Next