Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Oct 20, 2025

• Revert "[inductor] require shape in TritonCSEVariable (#162275)"

  Show description for 240c133
  pytorchmergebot

  committed
  240c133

  Copy full SHA for 240c133

• Revert "Remove workaround to old CUDA bug (#164354)"

  Show description for 150682b
  pytorchmergebot

  committed
  150682b

  Copy full SHA for 150682b

• Revert "Move toString(ScalarType) and ScalarType ostream operator to headeronly (#164405)"

  Show description for ca7360e
  pytorchmergebot

  committed
  ca7360e

  Copy full SHA for ca7360e

• Revert "[dynamo][user_defined] Replace UserFunctionVariable with VariableTracker build (#165706)"

  Show description for 0bf6043
  pytorchmergebot

  committed
  0bf6043

  Copy full SHA for 0bf6043

• Revert "[dynamo][misc] Replace UserFunctionVariable with VariableTracker build (#165707)"

  Show description for 9875e70
  pytorchmergebot

  committed
  9875e70

  Copy full SHA for 9875e70

• Revert "Refactor out headeronly ArrayRef (#164991)"

  Show description for 69a4bfe
  pytorchmergebot

  committed
  69a4bfe

  Copy full SHA for 69a4bfe

• Revert "Widen ops support to take in IntHOArrayRef vs only std::vec (#165152)"

  Show description for 62a263b
  pytorchmergebot

  committed
  62a263b

  Copy full SHA for 62a263b

• Revert "[Submodule] Bump FBGEMM to latest (#165544)"

  Show description for 0da1f91
  pytorchmergebot

  committed
  0da1f91

  Copy full SHA for 0da1f91

• [ROCm][CI] Update rocm.yml workflow to use 1 GPU ARC runners (#165481)

  Show description for 8700d68
  

  2 peopleauthored and

  pytorchmergebot

  committed
  8700d68

  Copy full SHA for 8700d68

• Revert "[1/N] Change C-style casts to static_cast or reinterpret_cast (#165750)"

  Show description for ab82456
  pytorchmergebot

  committed
  ab82456

  Copy full SHA for ab82456

• [Inductor][CuTeDSL] Move load_template up two directories (#165868)

  Show description for b23f468
  

  NikhilAPatel

  authored and

  pytorchmergebot

  committed
  b23f468

  Copy full SHA for b23f468

• [CI] Add rocm CI back to trunk for pre-submit/PR jobs (#165674)

  Show description for 2705937
  

  jithunnair-amd

  authored and

  pytorchmergebot

  committed
  2705937

  Copy full SHA for 2705937

• [cuda] fix triu/tril int32 overflow for large matrices (#164705)

  Show description for c1eda34
  

  Aminsed

  authored and

  pytorchmergebot

  committed
  c1eda34

  Copy full SHA for c1eda34

• [cuda] fix nll_loss2d backward bounds check with reduction=none (#165247)

  Show description for ba93d56
  

  Aminsed

  authored and

  pytorchmergebot

  committed
  ba93d56

  Copy full SHA for ba93d56

• [dynamo] Remove duplicated guards (#165806)

  Show description for 722b2b8
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  722b2b8

  Copy full SHA for 722b2b8

• [1/N] Change C-style casts to static_cast or reinterpret_cast (#165750)

  Show description for e1e8491
  

  cyyever

  authored and

  pytorchmergebot

  committed
  e1e8491

  Copy full SHA for e1e8491

• [flex_attention] replace sliced BlockMask noop with helpful error (#164702)

  Show description for 767199f
  

  Aminsed

  authored and

  pytorchmergebot

  committed
  767199f

  Copy full SHA for 767199f

• Revert "[ATen] Fix CUDA reduction warp shuffle order (#164790)"

  Show description for 602ace5
  pytorchmergebot

  committed
  602ace5

  Copy full SHA for 602ace5

• Revert "12/n : Remove fbandroid_compiler_flags (#165558)"

  Show description for 47804ce
  pytorchmergebot

  committed
  47804ce

  Copy full SHA for 47804ce

• [Inductor] support masked vectorization for the tail_loop for fp8 datatype (#163324)

  Show description for e8cb34d
  

  jiayisunx

  authored and

  pytorchmergebot

  committed
  e8cb34d

  Copy full SHA for e8cb34d

• [Inductor] support masked vectorization for the tail_loop for float64 datatype (#163316)

  Show description for e9d8973
  

  jiayisunx

  authored and

  pytorchmergebot

  committed
  e9d8973

  Copy full SHA for e9d8973

Commits on Oct 19, 2025

• [Fix XPU CI] [Inductor UT] Fix test cases broken by community. (#165714)

  Show description for 61d9a51
  

  etaf

  authored and

  pytorchmergebot

  committed
  61d9a51

  Copy full SHA for 61d9a51

• [ATen] Switch order of blocked reduce when vectorize loads (#165178)

  Show description for 8a8329b
  

  PaulZhang12

  authored and

  pytorchmergebot

  committed
  8a8329b

  Copy full SHA for 8a8329b

• [FlexAttention] Fix dynamic shaped heads flex_flash check (#165866)

  Show description for 6b80c94
  

  drisspg

  authored and

  pytorchmergebot

  committed
  6b80c94

  Copy full SHA for 6b80c94

• test_scaled_matmul_cuda: fix infer_scale_swizzle (#165788)

  Show description for 8951df0
  

  jagadish-amd

  authored and

  pytorchmergebot

  committed
  8951df0

  Copy full SHA for 8951df0

• [dynamo] Add recompile reason for set_stance fail_on_recompile (#165445)

  Show description for 8139f33
  

  parsshar-RH

  authored and

  pytorchmergebot

  committed
  8139f33

  Copy full SHA for 8139f33

• [dynamo] Clean up assert in dynamo [1/N] (#165430)

  Show description for a885873
  

  2 peopleauthored and

  pytorchmergebot

  committed
  a885873

  Copy full SHA for a885873

• Revert "shrink_group implementation to expose ncclCommShrink API (#164518)"

  Show description for 633a3b7
  pytorchmergebot

  committed
  633a3b7

  Copy full SHA for 633a3b7

• shrink_group implementation to expose ncclCommShrink API (#164518)

  Show description for fa0db21
  

  brchang24

  authored and

  pytorchmergebot

  committed
  fa0db21

  Copy full SHA for fa0db21

• Remove E721 suppression in flake8 (#165855)

  Show description for 15ff1cd
  

  cyyever

  authored and

  pytorchmergebot

  committed
  15ff1cd

  Copy full SHA for 15ff1cd

• Migrating some more callsites (#163580)

  Show description for c73f508
  

  tugsbayasgalan

  authored and

  pytorchmergebot

  committed
  c73f508

  Copy full SHA for c73f508

• AOTI util deprecated flow using the new tracer (#165582)

  Show description for 22ae059
  

  tugsbayasgalan

  authored and

  pytorchmergebot

  committed
  22ae059

  Copy full SHA for 22ae059

• Fix AllocatorConfig parse roundup division bug (#165304)

  Show description for 1b121d6
  

  guangyey

  authored and

  pytorchmergebot

  committed
  1b121d6

  Copy full SHA for 1b121d6

• Refine CUDA BackendStaticInitializer for allocator select (#165298)

  Show description for 1ba808d
  

  guangyey

  authored and

  pytorchmergebot

  committed
  1ba808d

  Copy full SHA for 1ba808d

• Introduce a generic API torch._C._accelerator_setAllocatorSettings (#165291)

  Show description for b2f5c25
  

  guangyey

  authored and

  pytorchmergebot

  committed
  b2f5c25

  Copy full SHA for b2f5c25

Pagination

Previous

Next