Commits

Branch selector

main

User selector

All users

All time

Commit history

Commits on Mar 23, 2026

• [ROCm] Reland: Enable expandable segments (#173330) (#177974)

  Show description for 5792701
  

  haoyuz

  authored and

  pytorchmergebot

  committed
  5792701

  Copy full SHA for 5792701

• Revert "[dynamo] Adjust more tests to prepare for inline_inbuilt_nn_module deprecation (#178067)"

  Show description for 1d3abfe
  pytorchmergebot

  committed
  1d3abfe

  Copy full SHA for 1d3abfe

• Add GEMM configs to XPU autotuning heuristic (#177647)

  Show description for 11d35a3
  

  wdziurdz

  authored and

  pytorchmergebot

  committed
  11d35a3

  Copy full SHA for 11d35a3

• [Inductor] repeat_interleave fx graph runnable fix (#177909) (#177909)

  Show description for 7b45702
  

  PaulZhang12

  authored and

  pytorchmergebot

  committed
  7b45702

  Copy full SHA for 7b45702

• [XPU] Enable skipped inductor test on Intel GPU - generalize 21 inductor test files and enable xpu (#174058)

  Show description for 6ff7923
  

  daisyden

  authored and

  pytorchmergebot

  committed
  6ff7923

  Copy full SHA for 6ff7923

• [XPU] Enable skipped inductor test on Intel GPU - generalize 12 inductor test files and enable xpu (#174057)

  Show description for 1884117
  

  daisyden

  authored and

  pytorchmergebot

  committed
  1884117

  Copy full SHA for 1884117

• [XPU] Enable skipped inductor test on Intel GPU - generalize code and enable xpu in efficient attention and flash attention test (#174056)

  Show description for ed60e56
  

  daisyden

  authored and

  pytorchmergebot

  committed
  ed60e56

  Copy full SHA for ed60e56

• F.interpolate cleanups - again (#177877)

  Show description for 1a4108c
  

  NicolasHug

  authored and

  pytorchmergebot

  committed
  1a4108c

  Copy full SHA for 1a4108c

• [Triton] [Inductor] Add non-TMA persistent MM Triton template for max-autotune (#177781) (#177781)

  Show description for 28f2520
  

  CRobeck

  authored and

  pytorchmergebot

  committed
  28f2520

  Copy full SHA for 28f2520

• [dynamo] Add end-to-end test for Event.synchronize() with D2H copies (#177615)

  Show description for d90ad8e
  

  mlazos

  authored and

  pytorchmergebot

  committed
  d90ad8e

  Copy full SHA for d90ad8e

• [inductor] Add lowering and control_deps support for synchronize_event (#177614)

  Show description for 7040cd2
  

  mlazos

  authored and

  pytorchmergebot

  committed
  7040cd2

  Copy full SHA for 7040cd2

• [dynamo] Add synchronize_event custom op for Event.synchronize() (#177613)

  Show description for 0aae4e3
  

  mlazos

  authored and

  pytorchmergebot

  committed
  0aae4e3

  Copy full SHA for 0aae4e3

• [user-streams] fix bug where certain APIs don't take cuda events (#178027)

  Show description for 7fe0d46
  

  mlazos

  authored and

  pytorchmergebot

  committed
  7fe0d46

  Copy full SHA for 7fe0d46

Commits on Mar 22, 2026

• [Inductor] Fix benchmark_example_value losing dtype on view unwrap (#176859)

  Show description for 3f14378
  

  NikhilAPatel

  authored and

  pytorchmergebot

  committed
  3f14378

  Copy full SHA for 3f14378

• [Inductor][NVGEMM] Refactor tests (#176847)

  Show description for e50dd94
  

  NikhilAPatel

  authored and

  pytorchmergebot

  committed
  e50dd94

  Copy full SHA for e50dd94

• [Inductor][NVGEMM] Drop tile_k from nvMatmulHeuristics matching (#176845)

  Show description for c20451d
  

  NikhilAPatel

  authored and

  pytorchmergebot

  committed
  c20451d

  Copy full SHA for c20451d

• [OSDC] Update setup-linux to support OSDC ARC runners (#177950)

  Show description for 613e2cf
  

  huydhn

  authored and

  pytorchmergebot

  committed
  613e2cf

  Copy full SHA for 613e2cf

• [xpu][fix]Fix -Wextra-semi warning by removing redundant semicolons (#177944)

  Show description for 74fb561
  

  chunhuanMeng

  authored and

  pytorchmergebot

  committed
  74fb561

  Copy full SHA for 74fb561

• [BE][Ez]: Add missing std::move on std::make_tuple return calls (#177982)

  Show description for 9b48434
  

  Skylion007

  authored and

  pytorchmergebot

  committed
  9b48434

  Copy full SHA for 9b48434

• Revert "[XPU] Enable skipped inductor test on Intel GPU - generalize code and enable xpu in efficient attention and flash attention test (#174056)"

  Show description for ecea083
  pytorchmergebot

  committed
  ecea083

  Copy full SHA for ecea083

• Revert "[XPU] Enable skipped inductor test on Intel GPU - generalize 12 inductor test files and enable xpu (#174057)"

  Show description for f1f30bc
  pytorchmergebot

  committed
  f1f30bc

  Copy full SHA for f1f30bc

• Revert "[XPU] Enable skipped inductor test on Intel GPU - generalize 21 inductor test files and enable xpu (#174058)"

  Show description for 7ea2f65
  pytorchmergebot

  committed
  7ea2f65

  Copy full SHA for 7ea2f65

• Use aminmax instead of min and max kernels in histc (#178011)

  Show description for 8b905db
  

  lakshayg

  authored and

  pytorchmergebot

  committed
  8b905db

  Copy full SHA for 8b905db

• [XPU] Enable skipped inductor test on Intel GPU - generalize 21 inductor test files and enable xpu (#174058)

  Show description for 7e68d42
  

  daisyden

  authored and

  pytorchmergebot

  committed
  7e68d42

  Copy full SHA for 7e68d42

• [XPU] Enable skipped inductor test on Intel GPU - generalize 12 inductor test files and enable xpu (#174057)

  Show description for d698aea
  

  daisyden

  authored and

  pytorchmergebot

  committed
  d698aea

  Copy full SHA for d698aea

• [XPU] Enable skipped inductor test on Intel GPU - generalize code and enable xpu in efficient attention and flash attention test (#174056)

  Show description for 9613520
  

  daisyden

  authored and

  pytorchmergebot

  committed
  9613520

  Copy full SHA for 9613520

• [XPU] Enable skipped inductor test on Intel GPU - enable xpu in eager tests that are reused by inductor tests (#174055)

  Show description for 044c88c
  

  daisyden

  authored and

  pytorchmergebot

  committed
  044c88c

  Copy full SHA for 044c88c

• [XPU] Enable skipped inductor test on Intel GPU - enable xpu enough memory check (#174054)

  Show description for 9c7971d
  

  daisyden

  authored and

  pytorchmergebot

  committed
  9c7971d

  Copy full SHA for 9c7971d

• [XPU] Enable skipped inductor test on Intel GPU - generalize code and enable xpu for functions under torch/ folder (#174053)

  Show description for af0ae02
  

  daisyden

  authored and

  pytorchmergebot

  committed
  af0ae02

  Copy full SHA for af0ae02

• [pytree][dynamo] refactor test_tree_map.py to use module-based pytree parametrization (#178050)

  Show description for 5ad794d
  

  XuehaiPan

  authored and

  pytorchmergebot

  committed
  5ad794d

  Copy full SHA for 5ad794d

• Revert "[MPS] Add API to load pre-compiled metallib (#177276)"

  Show description for 67f1ccf
  pytorchmergebot

  committed
  67f1ccf

  Copy full SHA for 67f1ccf

• [dynamo] Adjust more tests to prepare for inline_inbuilt_nn_module deprecation (#178067)

  Show description for 27c9a35
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  27c9a35

  Copy full SHA for 27c9a35

Commits on Mar 21, 2026

• [MPS] Add API to load pre-compiled metallib (#177276)

  Show description for fa15f2b
  

  imperatormk

  authored and

  pytorchmergebot

  committed
  fa15f2b

  Copy full SHA for fa15f2b

• Preserve cumulative out dtype semantics during functionalization (#177770)

  Show description for 3e6de85
  

  bobrenjc93

  authored and

  pytorchmergebot

  committed
  3e6de85

  Copy full SHA for 3e6de85

• Revert "Replace deprecated check in inductor diagonal lowering (#176930)"

  Show description for 06858a1
  pytorchmergebot

  committed
  06858a1

  Copy full SHA for 06858a1