Commits

Branch selector

 e2c7ae5

User selector

All users

Datepicker

All time

Commit History

Commits on Apr 25, 2025

• [ONNX] Add group_norm support from opset 21 (#152138)

  Show description for e2c7ae5
  

  justinchuby

  authored and

  pytorchmergebot

  committed
  e2c7ae5

  Copy full SHA for e2c7ae5

• Reducer: add check on received data to avoid segfault (#152143)

  Show description for 1a6d50d
  

  d4l3k

  authored and

  pytorchmergebot

  committed
  1a6d50d

  Copy full SHA for 1a6d50d

• Adding fbgemm to whitelist (#152079)

  Show description for 7f28c03
  

  jimone1

  authored and

  pytorchmergebot

  committed
  7f28c03

  Copy full SHA for 7f28c03

• Revert "Add OIDC permissions to bazel workflow (#151456)"

  Show description for 8313bc2
  pytorchmergebot

  committed
  8313bc2

  Copy full SHA for 8313bc2

• [Break XPU] generalize newly introduced device bias code in Inductor UT. (#151926)

  Show description for 75c71ab
  

  etaf

  authored and

  pytorchmergebot

  committed
  75c71ab

  Copy full SHA for 75c71ab

Commits on Apr 24, 2025

• [Inductor][CPP] Optimize the epilogue for int8 GEMM Template (#152000)

  Show description for d70490e
  

  leslie-fang-intel

  authored and

  pytorchmergebot

  committed
  d70490e

  Copy full SHA for d70490e

• [xpu] set aot device flags in cpp_extension (#149459)

  Show description for 2089b22
  

  3 peopleauthored and

  pytorchmergebot

  committed
  2089b22

  Copy full SHA for 2089b22

• [Inductor] Record Triton’s Base32 Cache Key in .best_config for Debugging (#148981)

  Show description for fc6e37c
  

  fulvius31

  authored and

  pytorchmergebot

  committed
  fc6e37c

  Copy full SHA for fc6e37c

• Non-deterministic alert in histc_cuda for floating types only (#151701)

  Show description for 0413358
  

  amjames

  authored and

  pytorchmergebot

  committed
  0413358

  Copy full SHA for 0413358

• Python 3.11 and 3.13 support for Windows Arm64 (#152109)

  Show description for 6ced5e6
  

  iremyux

  authored and

  pytorchmergebot

  committed
  6ced5e6

  Copy full SHA for 6ced5e6

• [CUDA][TF32] Account for TF32 in `test_corrcoef` (#151830)

  Show description for d78d2af
  

  eqy

  authored and

  pytorchmergebot

  committed
  d78d2af

  Copy full SHA for d78d2af

• Improve stable library apis per Scott's feedback (#152040)

  Show description for 8a9c66b
  

  janeyx99

  authored and

  pytorchmergebot

  committed
  8a9c66b

  Copy full SHA for 8a9c66b

• Include other accelerators in capturable docstr for optimizers (#149770)

  Show description for dccc415
  

  janeyx99

  authored and

  pytorchmergebot

  committed
  dccc415

  Copy full SHA for dccc415

• add Out Notes (#151306)

  Show description for bd09d87
  

  ILCSFNO

  authored and

  pytorchmergebot

  committed
  bd09d87

  Copy full SHA for bd09d87

• [export] improve error message for deserializing custom triton op (#152029)

  Show description for 92f125e
  

  ydwu4

  authored and

  pytorchmergebot

  committed
  92f125e

  Copy full SHA for 92f125e

• Pin theme to a branch (#152046)

  Show description for 24bda01
  

  svekars

  authored and

  pytorchmergebot

  committed
  24bda01

  Copy full SHA for 24bda01

• [CUDA][CPU] Bump system memory requirement for `test_cross_entropy_large_tensor` (#151812)

  Show description for 6efc572
  

  eqy

  authored and

  pytorchmergebot

  committed
  6efc572

  Copy full SHA for 6efc572

• Revert "[dynamo] Add guard serialization for tensor matches. (#151318)"

  Show description for b1d055f
  pytorchmergebot

  committed
  b1d055f

  Copy full SHA for b1d055f

• [CI][docker] Use install_cusparselt when possible in docker image (#150600)

  Show description for b11c9e1
  

  clee2000

  authored and

  pytorchmergebot

  committed
  b11c9e1

  Copy full SHA for b11c9e1

• Update docs dependencies for local build (#151796)

  Show description for ff075d0
  

  svekars

  authored and

  pytorchmergebot

  committed
  ff075d0

  Copy full SHA for ff075d0

• [dynamo] Add guard serialization for tensor matches. (#151318)

  Show description for 81c4369
  

  zhxchen17

  authored and

  pytorchmergebot

  committed
  81c4369

  Copy full SHA for 81c4369

• Add functionality for installing free variables (#151134)

  Show description for 03970df
  

  Lucaskabela

  authored and

  pytorchmergebot

  committed
  03970df

  Copy full SHA for 03970df

• add basic unit tests and noop config (#152036)

  Show description for 402d19c
  

  Lucaskabela

  authored and

  pytorchmergebot

  committed
  402d19c

  Copy full SHA for 402d19c

• [fake tensor] Cache None, integer and SymInts in the output (#151961)

  Show description for 9c1bc9c
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  9c1bc9c

  Copy full SHA for 9c1bc9c

• Better error msg for too big to optimize (#151855)

  Show description for 0eb554e
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  0eb554e

  Copy full SHA for 0eb554e

• Move verbose warning to warning_once (#152044)

  Show description for 56e67ba
  

  wconstab

  authored and

  pytorchmergebot

  committed
  56e67ba

  Copy full SHA for 56e67ba

• Revert "[Cutlass] Implement EVT example tensor creation (#150904)"

  Show description for 3a170a8
  pytorchmergebot

  committed
  3a170a8

  Copy full SHA for 3a170a8

• [invoke_subgraph] Cache fake tensor if no unbacked symint in the output (#151957)

  Show description for d743a7b
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  d743a7b

  Copy full SHA for d743a7b

• [fake tensor cache] Support index with non bool/int8 indices (#151477)

  Show description for 1d73b64
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  1d73b64

  Copy full SHA for 1d73b64

• [invoke_subgraph][fake tensor] Add finalizer on subgraph instead of the functionalize ctx wrapper (#151633)

  Show description for 41285f2
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  41285f2

  Copy full SHA for 41285f2

• [invoke_subgraph] Compile time traces (#151409)

  Show description for 3278ddd
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  3278ddd

  Copy full SHA for 3278ddd

• [BE] follow autoformating and linter (#151507)

  Show description for 5e320ee
  

  XilunWu

  authored and

  pytorchmergebot

  committed
  5e320ee

  Copy full SHA for 5e320ee

• Add torch.cuda._compile_kernel() (#151484)

  Show description for 5b368fa
  

  msaroufim

  authored and

  pytorchmergebot

  committed
  5b368fa

  Copy full SHA for 5b368fa

• [pytorch] reland of [cutlass backend] delay construction of cutlass presets to when called (#151875) (#152031)

  Show description for 78953ee
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  78953ee

  Copy full SHA for 78953ee

• [vec128] Fix fmsub NEON defintion (#152075)

  Show description for 2ea8653
  

  malfet

  authored and

  pytorchmergebot

  committed
  2ea8653

  Copy full SHA for 2ea8653

Pagination

Previous

Next