Commits

Branch selector

 main

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 2, 2025

• Clear custom autograd Function ctx.to_save earlier (#161171)

  Show description for 8171d60
  

  soulitzer

  authored and

  pytorchmergebot

  committed
  8171d60

  Copy full SHA for 8171d60

• Fixes broken memory_viz link in CUDA memory docs (#161426)

  Show description for d5e0f42
  

  dsashidh

  authored and

  pytorchmergebot

  committed
  d5e0f42

  Copy full SHA for d5e0f42

• [BE][Easy] restore #157584 after #158288 (#158541)

  Show description for 13d66e2
  

  XuehaiPan

  authored and

  pytorchmergebot

  committed
  13d66e2

  Copy full SHA for 13d66e2

• test: ensure editable cached wrapper is respected (#160943)

  Show description for bbedc71
  

  ezyang

  authored and

  pytorchmergebot

  committed
  bbedc71

  Copy full SHA for bbedc71

• [dynamo] Prevent unnecessary recompile on disabled functions in the compiled frame (#161883)

  Show description for e9481b6
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  e9481b6

  Copy full SHA for e9481b6

• Fix slice scatter dtype consistency (#160851)

  Show description for 1c1b28d
  

  ghostspiders

  authored and

  pytorchmergebot

  committed
  1c1b28d

  Copy full SHA for 1c1b28d

• [AOTI] split too long string to smaller pieces when its length larger than 16000, fix msvc c2026. (#161850)

  Show description for 2a5c078
  

  xuhancn

  authored and

  pytorchmergebot

  committed
  2a5c078

  Copy full SHA for 2a5c078

Commits on Sep 1, 2025

• Make distributed modules importable even when backend not built (#159889)

  Show description for 626cb7d
  

  ezyang

  authored and

  pytorchmergebot

  committed
  626cb7d

  Copy full SHA for 626cb7d

• Always build USE_DISTRIBUTED. (#160449)

  Show description for b7034e9
  

  ezyang

  authored and

  pytorchmergebot

  committed
  b7034e9

  Copy full SHA for b7034e9

• Revert "Defer loading hipify until it is needed (#160824)"

  Show description for 13b6519
  pytorchmergebot

  committed
  13b6519

  Copy full SHA for 13b6519

• Defer loading hipify until it is needed (#160824)

  Show description for 403a3a3
  

  guilhermeleobas

  authored and

  pytorchmergebot

  committed
  403a3a3

  Copy full SHA for 403a3a3

• Fix type checking for persistent loads in the weights-only unpickler (#161661)

  Show description for cbfb005
  

  dfyz

  authored and

  pytorchmergebot

  committed
  cbfb005

  Copy full SHA for cbfb005

• [BE] Consolidate inductor benchmark Docker images and rename jobs (#161536)

  Show description for d232a95
  

  huydhn

  authored and

  pytorchmergebot

  committed
  d232a95

  Copy full SHA for d232a95

• Revert "Fix conv exhaustive autotuning and expand Exhaustive test coverage (#159387)"

  Show description for 17fa8ee
  pytorchmergebot

  committed
  17fa8ee

  Copy full SHA for 17fa8ee

• Revert "[Fix XPU CI][Inductor UT] Fix test cases broken by community. (#161142)"

  Show description for 54e275e
  pytorchmergebot

  committed
  54e275e

  Copy full SHA for 54e275e

• Revert "[CUDA] Reuse blocks with record_stream during CUDA Graph capture in the CUDACachingAllocator (#158352)"

  Show description for 63a9c23
  pytorchmergebot

  committed
  63a9c23

  Copy full SHA for 63a9c23

• [CD] Add CUDA 13.0 Windows build (#161663)

  Show description for fefee08
  

  tinglvv

  authored and

  pytorchmergebot

  committed
  fefee08

  Copy full SHA for fefee08

• Revert "[cuBLASLt][FP8] `cuBLASLt` appears to support float8 rowwise-scaling on H100 (#161305)"

  Show description for 21fae99
  pytorchmergebot

  committed
  21fae99

  Copy full SHA for 21fae99

• [xla hash update] update the pinned xla hash (#161396)

  Show description for 2ba6547
  

  pytorchupdatebot

  authored and

  pytorchmergebot

  committed
  2ba6547

  Copy full SHA for 2ba6547

• [CUDA] Reuse blocks with record_stream during CUDA Graph capture in the CUDACachingAllocator (#158352)

  Show description for 190c391
  

  2 peopleauthored and

  pytorchmergebot

  committed
  190c391

  Copy full SHA for 190c391

• Skip compilation when FX graph has no calls and returns empty (#160536)

  Show description for 20bfb25
  

  Raman-RH

  authored and

  pytorchmergebot

  committed
  20bfb25

  Copy full SHA for 20bfb25

• ci: Update sphinx, disable google search by default (#161793)

  Show description for dd2519a
  

  seemethere

  authored and

  pytorchmergebot

  committed
  dd2519a

  Copy full SHA for dd2519a

• [4/N][SymmMem] Add `get_remote_tensor` + move up `get_buffer` and `get_signal_pad` (#161533)

  Show description for 2f6b4b1
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  2f6b4b1

  Copy full SHA for 2f6b4b1

• update supported OS for Intel client GPU (#161699)

  Show description for 6737e2c
  

  ZhaoqiongZ

  authored and

  pytorchmergebot

  committed
  6737e2c

  Copy full SHA for 6737e2c

• [vllm hash update] update the pinned vllm hash (#161867)

  Show description for 67c31dc
  

  pytorchupdatebot

  authored and

  pytorchmergebot

  committed
  67c31dc

  Copy full SHA for 67c31dc

• Remove background thread UT on XPU to fix CI (#161844)

  Show description for cb1e313
  

  guangyey

  authored and

  pytorchmergebot

  committed
  cb1e313

  Copy full SHA for cb1e313

Commits on Aug 31, 2025

• Add __init__.pyi to torch/linalg (#160750)

  Show description for 9a665ca
  

  stmcgovern

  authored and

  pytorchmergebot

  committed
  9a665ca

  Copy full SHA for 9a665ca

• Leak Python filenames so that we can give good dispatcher errors. (#160418)

  Show description for d9d6dde
  

  ezyang

  authored and

  pytorchmergebot

  committed
  d9d6dde

  Copy full SHA for d9d6dde

• PythonArgs::toBool: order cheap mutually exclusive checks first (#161455)

  Show description for 68738be
  

  swolchok

  authored and

  pytorchmergebot

  committed
  68738be

  Copy full SHA for 68738be

• [3/N][SymmMem] Expose offset field from handle (#161532)

  Show description for 25f4aae
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  25f4aae

  Copy full SHA for 25f4aae

• [2/N][SymmMem] Add MemPool allocator and tests (#161471)

  Show description for 61e18b5
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  61e18b5

  Copy full SHA for 61e18b5

• removed duplicate imports (#161685)

  Show description for e92cd94
  

  rohit-kumar-manav

  authored and

  pytorchmergebot

  committed
  e92cd94

  Copy full SHA for e92cd94

• fix spelling of word - when (#160185)

  Show description for 0d421ac
  

  Raman-RH

  authored and

  pytorchmergebot

  committed
  0d421ac

  Copy full SHA for 0d421ac

• [c10d][nvshmem] add nvshmem build rules and dependency for libtorch_cuda (#159562)

  Show description for 91f0bcf
  

  jushg

  authored and

  pytorchmergebot

  committed
  91f0bcf

  Copy full SHA for 91f0bcf

• [CPU][Inductor] Improve performance of A16W8 GEMM template (#161148)

  Show description for 75bc23c
  

  Xia-Weiwen

  authored and

  pytorchmergebot

  committed
  75bc23c

  Copy full SHA for 75bc23c

Pagination

Previous

Next