Commits

Breadcrumbs

History for

onmain

User selector

All users

Datepicker

All time

Commit History

Commits on Sep 18, 2024

• [BE][MPS] Fix deprecation warnings on MacOS 15.0 (pytorch#136292)

  

  malfet

  authored and

  pytorchmergebot

  committed
  068c80e
  More actions

• [BE][MPS] Delete duplicated code in View.mm (pytorch#136295)

  

  malfet

  authored and

  pytorchmergebot

  committed
  b9a197d
  More actions

• Extending the Pytorch vec backend for SVE (ARM) (pytorch#119571)

  

  2 peopleauthored and

  pytorchmergebot

  committed
  5a6ddbc
  More actions

• Add _addmm_activation to lower precision cast policy on AutocastCPU (pytorch#135936)

  

  CaoE

  authored and

  pytorchmergebot

  committed
  6a6f5b2
  More actions

• [BE] Make NestedTensorTransformerFunctions.cu compilable without warnings (pytorch#136222)

  

  malfet

  authored and

  pytorchmergebot

  committed
  6682327
  More actions

Commits on Sep 17, 2024

• [MPS] Fix 5D reductions over negative dimentions (pytorch#136198)

  

  malfet

  authored and

  pytorchmergebot

  committed
  f6f1504
  More actions

• [PyTorch Pinned Allocator] Add support of background thread to process events (pytorch#135524)

  

  banitag1

  authored and

  pytorchmergebot

  committed
  a575ce0
  More actions

• Support rms_norm() for NJT (pytorch#135872)

  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  a838284
  More actions

• [MTIA] Support torch.cuda.get_device_capability equivalent API on MTIA (pytorch#135889)

  

  ttrung149

  authored and

  pytorchmergebot

  committed
  cc365fd
  More actions

• make view.dtype always return an alias (pytorch#136074)

  

  bdhirsh

  authored and

  pytorchmergebot

  committed
  dc82d27
  More actions

Commits on Sep 16, 2024

• Remove accidentally committed code (pytorch#136154)

  

  malfet

  authored and

  pytorchmergebot

  committed
  d3647d1
  More actions

• [BE][Ez]: Fix missing float16 coverage for adaptive_pool3d_cpu (pytorch#136091)

  

  Skylion007

  authored and

  pytorchmergebot

  committed
  23c0d26
  More actions

• [BE][Ez]: Add full half/bfloat16 dtype for unique and isin (pytorch#136114)

  

  Skylion007

  authored and

  pytorchmergebot

  committed
  b491e29
  More actions

Commits on Sep 14, 2024

• Add Half support for reflection and replication padding on CPU (pytorch#135931)

  

  CaoE

  authored and

  pytorchmergebot

  committed
  db393fb
  More actions

• [MPS] Add native im2col (pytorch#135706)

  

  malfet

  authored and

  pytorchmergebot

  committed
  1a67e2b
  More actions

• Revert "Use _amp_foreach_non_finite_check_and_unscale_ for CPU grads of ShardedGradScaler (pytorch#135232)"

  pytorchmergebot

  committed
  1786a17
  More actions

• Use _amp_foreach_non_finite_check_and_unscale_ for CPU grads of ShardedGradScaler (pytorch#135232)

  

  CaoE

  authored and

  pytorchmergebot

  committed
  51c5206
  More actions

• [BE] Use squeeze/unsqueeze in im2col (pytorch#136006)

  

  malfet

  authored and

  pytorchmergebot

  committed
  081c4a9
  More actions

Commits on Sep 13, 2024

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

  pytorchmergebot

  committed
  6ef49fe
  More actions

• [BE] Use C10_UNUSED (pytorch#135914)

  

  malfet

  authored and

  pytorchmergebot

  committed
  1c04cbf
  More actions

Commits on Sep 12, 2024

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

  

  fadara01

  authored and

  pytorchmergebot

  committed
  3d24313
  More actions

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

  

  guilhermeleobas

  authored and

  pytorchmergebot

  committed
  f032135
  More actions

• NJT <-> padded dense conversions (pytorch#128637)

  

  jbschlosser

  authored and

  pytorchmergebot

  committed
  525bec8
  More actions

• Fix build warnings for torch_python (pytorch#134981)

  

  cyyever

  authored and

  pytorchmergebot

  committed
  f5f1d0a
  More actions

• Avoid inserting extra transpose when the input to group norm is NHWC (pytorch#135575)

  

  hk3911

  authored and

  pytorchmergebot

  committed
  7922311
  More actions

Commits on Sep 11, 2024

• fix for fp16 (pytorch#134106)

  

  mayank31398

  authored and

  pytorchmergebot

  committed
  9a04cfb
  More actions

• Expand bitwise ops to unsigned types (pytorch#135525)

  

  malfet

  authored and

  pytorchmergebot

  committed
  e20ee39
  More actions

• [ROCm] Update to AOTriton 0.7b (pytorch#134498)

  

  xinyazhang

  authored and

  pytorchmergebot

  committed
  74fd1bf
  More actions

• ATen | Fix MPSCNNNeuron creation on Mac Catalyst. (pytorch#135595)

  

  nlutsenko

  authored and

  pytorchmergebot

  committed
  ce4d146
  More actions

• [cuDNN][SDPA] Support attn_bias in cuDNN (pytorch#130482)

  

  2 peopleauthored and

  pytorchmergebot

  committed
  3084b7b
  More actions

• [MPS] Add missing dispatch to rshift.Tensor (pytorch#135607)

  

  malfet

  authored and

  pytorchmergebot

  committed
  3bf6be4
  More actions

• Add boolean support in pack segments ops for both cpu and cuda impls (pytorch#132897) (pytorch#135620)

  

  hlu90

  authored and

  pytorchmergebot

  committed
  02dcb07
  More actions

Commits on Sep 10, 2024

• [ROCm] slow torch.sum optimization by increasing max_values_per_thread in reduce config (pytorch#135397)

  

  hongxiayang

  authored and

  pytorchmergebot

  committed
  eb38ee2
  More actions

• Add more logging to TunableOp validators (pytorch#135396)

  

  zixi-qi

  authored and

  pytorchmergebot

  committed
  c4b84a4
  More actions

• [MPS] Allow nan mean reduction in nll_loss (pytorch#135434)

  

  hvaara

  authored and

  pytorchmergebot

  committed
  23b1486
  More actions

Pagination

PreviousNext