Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 13, 2024

• [inductor] Remove the batch fusion passes from being a default (pytorch#135922)

  Show description for 7dc1788
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  7dc1788

  Copy full SHA for 7dc1788

• [Inductor UT] Generalize device-bias code in test_triton_kernels.py introduced in pytorch#135530 (pytorch#135656)

  Show description for 9fd54d7
  

  etaf

  authored and

  pytorchmergebot

  committed
  9fd54d7

  Copy full SHA for 9fd54d7

• [Inductor UT] Generalize inductor UT for intel GPU (Part 2) (pytorch#134556)

  Show description for b38be72
  

  hoshibara

  authored and

  pytorchmergebot

  committed
  b38be72

  Copy full SHA for b38be72

• [inductor] More fixes on the keys of constants and signature dictionaries (pytorch#135406)

  Show description for e54b559
  

  Jokeren

  authored and

  pytorchmergebot

  committed
  e54b559

  Copy full SHA for e54b559

• [DCP][DSD] Add a test case to demonstrate the workaround to load full state dict into a 2D model (pytorch#135763)

  Show description for eea5e6f
  

  wz337

  authored and

  pytorchmergebot

  committed
  eea5e6f

  Copy full SHA for eea5e6f

• real tensor prop for composite ops (pytorch#135717)

  Show description for 6df91b5
  

  pianpwk

  authored and

  pytorchmergebot

  committed
  6df91b5

  Copy full SHA for 6df91b5

• [DSD] Fix distributed state dict full_state_dict option hang during set_state_dict (pytorch#135725)

  Show description for 0cdc6a8
  

  wz337

  authored and

  pytorchmergebot

  committed
  0cdc6a8

  Copy full SHA for 0cdc6a8

• [ROCm] skip test_fp8_cast_and_t on non-MI300 machines (pytorch#135917)

  Show description for 6cdc70b
  

  pragupta

  authored and

  pytorchmergebot

  committed
  6cdc70b

  Copy full SHA for 6cdc70b

• Fix xpu memory stats error (pytorch#135818)

  Show description for e6b6835
  

  guangyey

  authored and

  pytorchmergebot

  committed
  e6b6835

  Copy full SHA for e6b6835

• [BE] Use C10_UNUSED (pytorch#135914)

  Show description for 1c04cbf
  

  malfet

  authored and

  pytorchmergebot

  committed
  1c04cbf

  Copy full SHA for 1c04cbf

• [Profiler] Torch Profiler distributed info is not JSON serializable (pytorch#135548)

  Show description for 062681a
  

  sraikund16

  authored and

  pytorchmergebot

  committed
  062681a

  Copy full SHA for 062681a

• Fix lint errors in fbcode (pytorch#135614)

  Show description for 8c356ce
  

  aorenste

  authored and

  pytorchmergebot

  committed
  8c356ce

  Copy full SHA for 8c356ce

• [dynamo] Fix support for classmethod(property(...)) (pytorch#134968)

  Show description for bf68e16
  

  jansel

  authored and

  pytorchmergebot

  committed
  bf68e16

  Copy full SHA for bf68e16

• [Inductor] Disable TF32 in test_slice_scatter_reinplace (pytorch#135709)

  Show description for d732df7
  

  eqy

  authored and

  pytorchmergebot

  committed
  d732df7

  Copy full SHA for d732df7

• [Docs] fix inconsistent docs in conv1d, conv2d, and conv3d (pytorch#135894)

  Show description for c9de2ef
  

  PaliC

  authored and

  pytorchmergebot

  committed
  c9de2ef

  Copy full SHA for c9de2ef

• [fx] Replace _snake_case with a regexp (pytorch#135822)

  Show description for 1f15c0c
  

  jansel

  authored and

  pytorchmergebot

  committed
  1f15c0c

  Copy full SHA for 1f15c0c

• [fx] Minor optimization in create_arg (pytorch#135821)

  Show description for a72124a
  

  jansel

  authored and

  pytorchmergebot

  committed
  a72124a

  Copy full SHA for a72124a

• [inductor] Use TracerBase directly in LoopBody (pytorch#135820)

  Show description for 10ca4c0
  

  jansel

  authored and

  pytorchmergebot

  committed
  10ca4c0

  Copy full SHA for 10ca4c0

• [inductor] Optimize can_fuse_vertical() (pytorch#135788)

  Show description for d3aab96
  

  jansel

  authored and

  pytorchmergebot

  committed
  d3aab96

  Copy full SHA for d3aab96

• [inductor] Remove unused check (pytorch#135787)

  Show description for 67a929e
  

  jansel

  authored and

  pytorchmergebot

  committed
  67a929e

  Copy full SHA for 67a929e

• do not expand in replace/simplify if no changes (pytorch#135863)

  Show description for f576960
  

  isuruf

  authored and

  pytorchmergebot

  committed
  f576960

  Copy full SHA for f576960

• Update nightly PyTorch version to 2.6.0 (pytorch#135916)

  Show description for 1aba224
  

  malfet

  authored and

  pytorchmergebot

  committed
  1aba224

  Copy full SHA for 1aba224

Commits on Sep 12, 2024

• [aoti] Fix workspace generation for triton (pytorch#135552)

  Show description for d383325
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  d383325

  Copy full SHA for d383325

• fix compiled_autograd deadlock throw (pytorch#135795)

  Show description for 00dc7d4
  

  majian4work

  authored and

  pytorchmergebot

  committed
  00dc7d4

  Copy full SHA for 00dc7d4

• [FlexAttention] Ensure q/k/v and block_mask on excact the same device (pytorch#135823)

  Show description for 1760bbc
  

  yanboliang

  authored and

  pytorchmergebot

  committed
  1760bbc

  Copy full SHA for 1760bbc

• [ROCm] Use ieee precision for fp32 in flex attention (pytorch#135702)

  Show description for fb9d8e3
  

  jataylo

  authored and

  pytorchmergebot

  committed
  fb9d8e3

  Copy full SHA for fb9d8e3

• [Easy] Check if quant registered in constant folding (pytorch#135875)

  Show description for aaabfc8
  

  eellison

  authored and

  pytorchmergebot

  committed
  aaabfc8

  Copy full SHA for aaabfc8

• [dynamo] support torch.nn.attention.sdpa_kernel context manager (pytorch#135404)

  Show description for 63d6cd3
  

  williamwen42

  authored and

  pytorchmergebot

  committed
  63d6cd3

  Copy full SHA for 63d6cd3

• Revert "Check function declarations of Core ML code (pytorch#135467)"

  Show description for 3de9e47
  pytorchmergebot

  committed
  3de9e47

  Copy full SHA for 3de9e47

• Revert "[DSD] Fix distributed state dict full_state_dict option hang during set_state_dict (pytorch#135725)"

  Show description for 3e1a4ea
  pytorchmergebot

  committed
  3e1a4ea

  Copy full SHA for 3e1a4ea

• Validate input types for torch.nn.Linear and torch.nn.Bilinear (pytorch#135596)

  Show description for e157ce3
  

  sanskarmodi8

  authored and

  pytorchmergebot

  committed
  e157ce3

  Copy full SHA for e157ce3

• [export] ignore mark_dynamic() in export (pytorch#135536)

  Show description for b897ab0
  

  pianpwk

  authored and

  pytorchmergebot

  committed
  b897ab0

  Copy full SHA for b897ab0

• Pass ideep:lowp_kind to matmul_forward::compute on cache misses (pytorch#135058)

  Show description for 3d24313
  

  fadara01

  authored and

  pytorchmergebot

  committed
  3d24313

  Copy full SHA for 3d24313

• [torch][fx] Add new replacement_callback to materialize a replacement just in time (pytorch#135553)

  Show description for cd472bb
  

  dulinriley

  authored and

  pytorchmergebot

  committed
  cd472bb

  Copy full SHA for cd472bb

• Add batching rule for torch.scatter_reduce (pytorch#135547)

  Show description for f032135
  

  guilhermeleobas

  authored and

  pytorchmergebot

  committed
  f032135

  Copy full SHA for f032135

Pagination

Previous

Next