Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on May 8, 2025

• Add torch._C.Tag.needs_contiguous_strides (#152859)

  Show description for 94ca3a4
  

  zou3519

  authored and

  pytorchmergebot

  committed
  94ca3a4

  Copy full SHA for 94ca3a4

• [1/n][Optimus][Auto-AC] Support activation quantization without scaling (#148380)

  Show description for 2d25e4d
  

  mengluy0125

  authored and

  pytorchmergebot

  committed
  2d25e4d

  Copy full SHA for 2d25e4d

• [dynamo] Fix bug in hasattr(tensor, "size") (#152883)

  Show description for 6f6fac6
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  6f6fac6

  Copy full SHA for 6f6fac6

• Add runtime asserts to AOTI (#152125)

  Show description for 834bc5e
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  834bc5e

  Copy full SHA for 834bc5e

Commits on May 7, 2025

• [Dynamo] Allow inlining into AO quantization modules (#152934)

  Show description for 20e2ca3
  

  mlazos

  authored and

  pytorchmergebot

  committed
  20e2ca3

  Copy full SHA for 20e2ca3

• [Cutlass] Handle broadcasting in EVT python codegen (#152733)

  Show description for 6b9d741
  

  mlazos

  authored and

  pytorchmergebot

  committed
  6b9d741

  Copy full SHA for 6b9d741

• Fix test/test_optim.py error message. (#153076)

  Show description for 4270517
  

  MeetThePatel

  authored and

  pytorchmergebot

  committed
  4270517

  Copy full SHA for 4270517

• [CUDA][cuDNN] Fix handling of `CPU` side input and target length tensors in `CTCLoss` (#152745)

  Show description for cecfc7d
  

  eqy

  authored and

  pytorchmergebot

  committed
  cecfc7d

  Copy full SHA for cecfc7d

• [ONNX] dynamic_shapes uses DYNAMIC (#153065)

  Show description for 773a91c
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  773a91c

  Copy full SHA for 773a91c

• [nativert] Move MPMCQueue to torch/nativert. (#152837)

  Show description for 5bb154e
  

  zhxchen17

  authored and

  pytorchmergebot

  committed
  5bb154e

  Copy full SHA for 5bb154e

• [Inductor] Set correct baseline for decomposek test (#152897)

  Show description for d2ee606
  

  PaulZhang12

  authored and

  pytorchmergebot

  committed
  d2ee606

  Copy full SHA for d2ee606

• [cutlass backend][test] re-enable test_cuda_compile_command for fbcode (#153001)

  Show description for dd7d231
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  dd7d231

  Copy full SHA for dd7d231

• [Dynamo] Guard serialization for RANGE_ITERATOR_MATCH (#152872)

  Show description for d9b8473
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  d9b8473

  Copy full SHA for d9b8473

• [Dynamo] Guard serialization for TUPLE_ITERATOR_LEN (#152865)

  Show description for 52f7106
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  52f7106

  Copy full SHA for 52f7106

• [Dynamo] Guard serialization for SEQUENCE_LENGTH (#152730)

  Show description for fb500d0
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  fb500d0

  Copy full SHA for fb500d0

• [Dynamo] Guard serialization for CLOSURE_MATCH (#152728)

  Show description for 42954ab
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  42954ab

  Copy full SHA for 42954ab

• [Dynamo] Guard serialization for FUNCTION_MATCH (#152727)

  Show description for a9186ec
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  a9186ec

  Copy full SHA for a9186ec

• [Dynamo] Guard serialization for NN_MODULE (#152725)

  Show description for a6f51be
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  a6f51be

  Copy full SHA for a6f51be

• [CUDA] Rest peak memory stats before running `test_set_per_process_memory_fraction` (#152540)

  Show description for 172e641
  

  eqy

  authored and

  pytorchmergebot

  committed
  172e641

  Copy full SHA for 172e641

• [cutlass backend] cache filtered ops based on layouts (#152580)

  Show description for 8b9c9a3
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  8b9c9a3

  Copy full SHA for 8b9c9a3

• [Break XPU] Fix XPU UT failures introduced by community. (#152945)

  Show description for 56879f6
  

  etaf

  authored and

  pytorchmergebot

  committed
  56879f6

  Copy full SHA for 56879f6

• Add infra to run CPython tests under Dynamo (#150787)

  Show description for ae1e51b
  

  guilhermeleobas

  authored and

  pytorchmergebot

  committed
  ae1e51b

  Copy full SHA for ae1e51b

• [dynamo] Recursively realize the stack_values (#152853)

  Show description for ecd74c9
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  ecd74c9

  Copy full SHA for ecd74c9

• [aoti] skip input symbol codegen for sympy expr w/ many symbols (#152579)

  Show description for 81b6920
  

  ColinPeppler

  authored and

  pytorchmergebot

  committed
  81b6920

  Copy full SHA for 81b6920

• Revert "[aot][ca] save bw_module in AOTAutogradCache (#151860)"

  Show description for a28dcdb
  pytorchmergebot

  committed
  a28dcdb

  Copy full SHA for a28dcdb

• Revert "[ca] mark scalar int sizes as dynamic via tensor wrapping (#151731)"

  Show description for f6db749
  pytorchmergebot

  committed
  f6db749

  Copy full SHA for f6db749

• Revert "[ca] hide unused scalar int sizes from dynamo (#151962)"

  Show description for 8f208dc
  pytorchmergebot

  committed
  8f208dc

  Copy full SHA for 8f208dc

• Revert "[dynamo][ca] support dynamic annotations on tensors in ListVariables/TupleVariables (#152119)"

  Show description for 64bbf58
  pytorchmergebot

  committed
  64bbf58

  Copy full SHA for 64bbf58

• [MPS] SDPA specialized kernels (#152781)

  Show description for 56492bf
  

  Isalia20

  authored and

  pytorchmergebot

  committed
  56492bf

  Copy full SHA for 56492bf

• [Dynamo] Guard serialization for CONSTANT_MATCH (#152724)

  Show description for 2b2b790
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  2b2b790

  Copy full SHA for 2b2b790

• [Dynamo] Guard serialization for EQUALS_MATCH (#152704)

  Show description for 6d1e899
  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  6d1e899

  Copy full SHA for 6d1e899

Commits on May 6, 2025

• [cutlass backend][BE][clean-up] refactor to remove use of autotune_fallback_to_aten=True in cutlass backend tests (#152850)

  Show description for 61aa77e
  

  henrylhtsang

  authored and

  pytorchmergebot

  committed
  61aa77e

  Copy full SHA for 61aa77e

• [ONNX] Suggest users setting dynamo=True when exporting (#152478)

  Show description for 5fa5017
  

  titaiwangms

  authored and

  pytorchmergebot

  committed
  5fa5017

  Copy full SHA for 5fa5017

• [dynamo] Support `delattr` on result of `torch.compile(module)` (#152741)

  Show description for 6c025b5
  

  StrongerXi

  authored and

  pytorchmergebot

  committed
  6c025b5

  Copy full SHA for 6c025b5

• [dynamo] Avoid running `torch.nn.Module.__call__` twice under `torch.compile(mod)` (#152740)

  Show description for 0886d40
  

  StrongerXi

  authored and

  pytorchmergebot

  committed
  0886d40

  Copy full SHA for 0886d40

Pagination

Previous

Next