Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on May 2, 2025

• [Inductor] Introduce Wrapper IR line for symbolic call args (#152587)

  Show description for add4702
  

  blaine-rister

  authored and

  pytorchmergebot

  committed
  add4702

  Copy full SHA for add4702

• [inductor] Realize bucketize/searchsorted output (#152644)

  Show description for 9ae4906
  

  davidberard98

  authored and

  pytorchmergebot

  committed
  9ae4906

  Copy full SHA for 9ae4906

• [DCP] Add 30min timeout for IPC communications in async checkpointing (#152629)

  Show description for a6dd1c2
  

  MeetVadakkanchery

  authored and

  pytorchmergebot

  committed
  a6dd1c2

  Copy full SHA for a6dd1c2

• Optimize `Sequential` methods description (#147304)

  Show description for d457b44
  

  zeshengzong

  authored and

  pytorchmergebot

  committed
  d457b44

  Copy full SHA for d457b44

• [CUDA][complex] skip `test_reference_numerics_large_jiterator_unary_cuda_complex64` on CUDA (#148024)

  Show description for 216d81d
  

  eqy

  authored and

  pytorchmergebot

  committed
  216d81d

  Copy full SHA for 216d81d

• [AOTAutogradCache][Easy] Move `"einops.einops.rearrange"` to `SAFE_NON_TORCH_FUNCTIONS` (#152640)

  Show description for 16153a0
  

  StrongerXi

  authored and

  pytorchmergebot

  committed
  16153a0

  Copy full SHA for 16153a0

• [ROCm] Add support for SymmetricMemory (#150580)

  Show description for 1ea2731
  

  2 peopleauthored and

  pytorchmergebot

  committed
  1ea2731

  Copy full SHA for 1ea2731

• consolidate guard_or_x and definitely_x (#152463)

  Show description for 376529c
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  376529c

  Copy full SHA for 376529c

• [StaticCudaLauncher] Ensure cuda context exists before launching kernels (#152667)

  Show description for 3baa85c
  

  jamesjwu

  authored and

  pytorchmergebot

  committed
  3baa85c

  Copy full SHA for 3baa85c

• Avoid triggering ignored requires_grad warning in our code (#152686)

  Show description for f51bee1
  

  albanD

  authored and

  pytorchmergebot

  committed
  f51bee1

  Copy full SHA for f51bee1

• [ROCm] Upgrade ROCm CI to ROCm6.4 (#151368)

  Show description for 844842d
  

  2 peopleauthored and

  pytorchmergebot

  committed
  844842d

  Copy full SHA for 844842d

• Make PGO code state not sensitive to file path by hashing file content when the file is available. (#152628)

  Show description for f65fb0a
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  f65fb0a

  Copy full SHA for f65fb0a

• [invoke_subgraph] Simplify output code for subgraph output node (#152490)

  Show description for ea4b7e0
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  ea4b7e0

  Copy full SHA for ea4b7e0

• [inductor][subgraph] Simplify the resulting output code for subgraph (#152383)

  Show description for f6761f2
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  f6761f2

  Copy full SHA for f6761f2

• Revert "[cutlass backend] Minor lru_cache to slightly speed up filtering ops (#152577)"

  Show description for 50d4698
  pytorchmergebot

  committed
  50d4698

  Copy full SHA for 50d4698

• Fix: Consider input defined unbacked during inductor codegen for runtime asserts (#152231)

  Show description for 38a9a8b
  

  laithsakka

  authored and

  pytorchmergebot

  committed
  38a9a8b

  Copy full SHA for 38a9a8b

• [SymmMem] Add all_to_all_vdev (#151819)

  Show description for 829752b
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  829752b

  Copy full SHA for 829752b

• [inductor][invoke_subgraph] Remove assertion checks for outputs of invoke_subgraph (#152384)

  Show description for 3731b70
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  3731b70

  Copy full SHA for 3731b70

• [invoke_subgraph] rename identifiers to prevent python mangling (#152581)

  Show description for 9e3fc41
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  9e3fc41

  Copy full SHA for 9e3fc41

• Revert "Use swap_tensors path in nn.Module.to for all subclasses that override __torch_dispatch__ (#152539)"

  Show description for 4f9f1ab
  pytorchmergebot

  committed
  4f9f1ab

  Copy full SHA for 4f9f1ab

• [SymmMem] Add all-to-all (#151498)

  Show description for d7961a1
  

  kwen2501

  authored and

  pytorchmergebot

  committed
  d7961a1

  Copy full SHA for d7961a1

• Revert "[Inductor] Add decomposeK as an autotuning choice for mm (#150654)"

  Show description for 7c3e679
  pytorchmergebot

  committed
  7c3e679

  Copy full SHA for 7c3e679

• [invoke_subgraph] Unpacked operands (#152547)

  Show description for 4649fd1
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  4649fd1

  Copy full SHA for 4649fd1

• Remove unnecessary __STDC_FORMAT_MACROS macro (#152513)

  Show description for ac5de6d
  

  FFFrog

  authored and

  pytorchmergebot

  committed
  ac5de6d

  Copy full SHA for ac5de6d

• [CUDAGraph Trees] support memory allocation on side stream (#152472)

  Show description for d969e2e
  

  BoyuanFeng

  authored and

  pytorchmergebot

  committed
  d969e2e

  Copy full SHA for d969e2e

• [ez] fix grammar mistakes in StatefulSymbolicContext comment (#152598)

  Show description for 1f89865
  

  bobrenjc93

  authored and

  pytorchmergebot

  committed
  1f89865

  Copy full SHA for 1f89865

• [CP] Fix the offsets to KV in backward (#152625)

  Show description for 36e5ff6
  

  fegin

  authored and

  pytorchmergebot

  committed
  36e5ff6

  Copy full SHA for 36e5ff6

• [cutlass backend] Minor lru_cache to slightly speed up filtering ops (#152577)

  Show description for 1fef3cd
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  1fef3cd

  Copy full SHA for 1fef3cd

• [refactor] refactor dense implementation of auto_functionalized_v2 for better clarity (#152248)

  Show description for 5b59389
  

  ydwu4

  authored and

  pytorchmergebot

  committed
  5b59389

  Copy full SHA for 5b59389

• [hop] make materialize_as_graph's include and exclude dispatch key set optional (#152247)

  Show description for 380327c
  

  ydwu4

  authored and

  pytorchmergebot

  committed
  380327c

  Copy full SHA for 380327c

• [hop][schema] allow adding kw_only info to schema argument (#152246)

  Show description for a776a56
  

  ydwu4

  authored and

  pytorchmergebot

  committed
  a776a56

  Copy full SHA for a776a56

• [hop][be] make check_input_alias_and_mutation_return_ouputs create new fake mode (#152245)

  Show description for 7e7b9ca
  

  ydwu4

  authored and

  pytorchmergebot

  committed
  7e7b9ca

  Copy full SHA for 7e7b9ca

• [Environment Variable][Rebase] Use thread-safe getenv functions (#140200)

  Show description for ce94b21
  

  cyyever

  authored and

  pytorchmergebot

  committed
  ce94b21

  Copy full SHA for ce94b21

• [ONNX] Delete JitTraceConvertStrategy (#152556)

  Show description for a5dd701
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  a5dd701

  Copy full SHA for a5dd701

• [inductor] if unbacked symint in old-size or new-size skip mark_reuse check (#152379)

  Show description for 3c54e0c
  

  ColinPeppler

  authored and

  pytorchmergebot

  committed
  3c54e0c

  Copy full SHA for 3c54e0c

Pagination

Previous

Next