Commits

Branch selector

 master

User selector

All users

Datepicker

All time

Commit History

Commits on Aug 16, 2024

• [Traceable FSDP2] Set torch._dynamo.config.skip_fsdp_hooks to True by default (pytorch#133531)

  Show description for 6790eb5
  

  yf225

  authored and

  pytorchmergebot

  committed
  6790eb5

  Copy full SHA for 6790eb5

• [Traceable FSDPS] Allow tracing through FSDP2 impl in trace_rules.py (pytorch#133532)

  Show description for 6d85077
  

  yf225

  authored and

  pytorchmergebot

  committed
  6d85077

  Copy full SHA for 6d85077

• S390x nightly binaries for python 3.13 (pytorch#132984)

  Show description for 18705e3
  

  AlekseiNikiforovIBM

  authored and

  pytorchmergebot

  committed
  18705e3

  Copy full SHA for 18705e3

• [Dynamo] Support torch.cuda.device ctx manager (pytorch#133385)

  Show description for 770086f
  

  yanboliang

  authored and

  pytorchmergebot

  committed
  770086f

  Copy full SHA for 770086f

• mixed_mm: add more extensive dtype testing (pytorch#133292)

  Show description for 38e5ee1
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  38e5ee1

  Copy full SHA for 38e5ee1

• [Profiler/CPU] Add API for Dynamic Activity Toggling [3/n] (pytorch#133353)

  Show description for 9c2d119
  

  sraikund16

  authored and

  pytorchmergebot

  committed
  9c2d119

  Copy full SHA for 9c2d119

• [c10d] Do not call ncclCommAbort if comm is not initialized (pytorch#133630)

  Show description for 46af996
  

  shuqiangzhang

  authored and

  pytorchmergebot

  committed
  46af996

  Copy full SHA for 46af996

• AutoHeuristic: documentation for mm (pytorch#133611)

  Show description for 8b8b4e5
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  8b8b4e5

  Copy full SHA for 8b8b4e5

• AutoHeuristic: mm ranking heuristic h100 (pytorch#133608)

  Show description for 0e0077f
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  0e0077f

  Copy full SHA for 0e0077f

• AutoHeuristic: Heuristic that ranks choices for mm (pytorch#131714)

  Show description for e51c8ad
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  e51c8ad

  Copy full SHA for e51c8ad

• [BE]: Update ruff to 0.6.0 (pytorch#133609)

  Show description for 51e1374
  

  Skylion007

  authored and

  pytorchmergebot

  committed
  51e1374

  Copy full SHA for 51e1374

• [inductor][cpp][gemm] fix k-slicing bug and add thread blocking config (pytorch#132730)

  Show description for eca8b42
  

  jgong5

  authored and

  pytorchmergebot

  committed
  eca8b42

  Copy full SHA for eca8b42

• Move python 3.8 to 3.9 for linux-binary-manywheel workflow (pytorch#133621)

  Show description for a6aa451
  

  atalman

  authored and

  pytorchmergebot

  committed
  a6aa451

  Copy full SHA for a6aa451

• Revert "[Flight Recorder] Add more basic analysis to the script (pytorch#133412)"

  Show description for e1b9b89
  pytorchmergebot

  committed
  e1b9b89

  Copy full SHA for e1b9b89

• Fix printing symfloat pow in triton (pytorch#133614)

  Show description for b444343
  

  isuruf

  authored and

  pytorchmergebot

  committed
  b444343

  Copy full SHA for b444343

• [inductor] [cpp] fix accuracy when template_buffer has users other than the epilogue nodes (pytorch#133073)

  Show description for 762b1b4
  

  chunyuan-w

  authored and

  pytorchmergebot

  committed
  762b1b4

  Copy full SHA for 762b1b4

• deprecate search_autotune_cache (pytorch#133628)

  Show description for dd69013
  

  nmacchioni

  authored and

  pytorchmergebot

  committed
  dd69013

  Copy full SHA for dd69013

• overestimate time_taken_ns for autotuning (pytorch#133633)

  Show description for 15183f5
  

  nmacchioni

  authored and

  pytorchmergebot

  committed
  15183f5

  Copy full SHA for 15183f5

• Remove AMD restrictions on triton hashing (pytorch#133616)

  Show description for 30fbf5b
  

  oulgen

  authored and

  pytorchmergebot

  committed
  30fbf5b

  Copy full SHA for 30fbf5b

• remove redundant upper bound check at runtime (pytorch#133627)

  Show description for 5ed3b70
  

  avikchaudhuri

  authored and

  pytorchmergebot

  committed
  5ed3b70

  Copy full SHA for 5ed3b70

• Update source matcher to use torch_fn (pytorch#133642)

  Show description for f64146a
  

  angelayi

  authored and

  pytorchmergebot

  committed
  f64146a

  Copy full SHA for f64146a

• Add auto-tuning for sparse semi-structured MM operator (pytorch#123742)

  Show description for d12bbcd
  

  alexsamardzic

  authored and

  pytorchmergebot

  committed
  d12bbcd

  Copy full SHA for d12bbcd

• [ROCm][CK][Inductor] enable dynamic shapes for CK backend to gemm max autotune (pytorch#133285)

  Show description for 3d45717
  

  tenpercent

  authored and

  pytorchmergebot

  committed
  3d45717

  Copy full SHA for 3d45717

• [PT2][Optimus] Add missing example value for the nodes introduced in group batch fusion (pytorch#133414)

  Show description for 8ea5b57
  

  mengluy0125

  authored and

  pytorchmergebot

  committed
  8ea5b57

  Copy full SHA for 8ea5b57

• [dynamo][user_defined][stable-diffusion] Raise ObservedAttributeError on UserDefinedObject var_getattr (pytorch#132806)

  Show description for 8a2b064
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  8a2b064

  Copy full SHA for 8a2b064

• [Flight Recorder] Add more basic analysis to the script (pytorch#133412)

  Show description for fcc2fc1
  

  fduwjj

  authored and

  pytorchmergebot

  committed
  fcc2fc1

  Copy full SHA for fcc2fc1

• [fx] Do not add Proxy on Tensor (pytorch#133470)

  Show description for d9f17cf
  

  yushangdi

  authored and

  pytorchmergebot

  committed
  d9f17cf

  Copy full SHA for d9f17cf

• [dynamo] Support object creation of classes with custom __new__ (pytorch#132977)

  Show description for 8a5708b
  

  anijain2305

  authored and

  pytorchmergebot

  committed
  8a5708b

  Copy full SHA for 8a5708b

• [ts_converter][reland] Add support for LinearOpContext and Conv2dOpContext in quantization pass (pytorch#133622)

  Show description for a1a869f
  

  angelayi

  authored and

  pytorchmergebot

  committed
  a1a869f

  Copy full SHA for a1a869f

• Fix type promotion for ldexp (pytorch#133519)

  Show description for 1653f77
  

  malfet

  authored and

  pytorchmergebot

  committed
  1653f77

  Copy full SHA for 1653f77

• AutoHeuristic: Enable explicit support for ranking (pytorch#131710)

  Show description for 3a904d1
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  3a904d1

  Copy full SHA for 3a904d1

• AutoHeuristic: Support ranking/pruning choices (pytorch#131705)

  Show description for add0f00
  

  AlnisM

  authored and

  pytorchmergebot

  committed
  add0f00

  Copy full SHA for add0f00

• [3/N] Fix clang-tidy warnings in torch/csrc/autograd (pytorch#133389)

  Show description for 929d2f8
  

  cyyever

  authored and

  pytorchmergebot

  committed
  929d2f8

  Copy full SHA for 929d2f8

• [inductor][cpp][gemm] improve large bs perf with better cache blocking (pytorch#132729)

  Show description for c22f51c
  

  jgong5

  authored and

  pytorchmergebot

  committed
  c22f51c

  Copy full SHA for c22f51c

• [14/N] Use std::optional (pytorch#133417)

  Show description for 8f7cf79
  

  cyyever

  authored and

  pytorchmergebot

  committed
  8f7cf79

  Copy full SHA for 8f7cf79

Pagination

Previous

Next