Commits

Branch selector

 master

User selector

t-vi

Datepicker

All time

Commit History

Commits on Mar 10, 2021

• Fix distributions which don't properly honor validate_args=False (#53600)

  Show description for 76b58dd
  

  t-vi

  authored and

  facebook-github-bot

  committed
  76b58dd

  Copy full SHA for 76b58dd

• Use native ctc loss for target length 256 (#53557)

  Show description for e13ef77
  

  t-vi

  authored and

  facebook-github-bot

  committed
  e13ef77

  Copy full SHA for e13ef77

Commits on Feb 10, 2021

• Keep alive graph when creating iterators from it (#51951)

  Show description for bd62481
  

  t-vi

  authored and

  facebook-github-bot

  committed
  bd62481

  Copy full SHA for bd62481

Commits on Feb 4, 2021

• Graceful invalidation of Python Node/Value/Block when C++ object is deleted (#50326)

  Show description for 8686109
  

  t-vi

  authored and

  facebook-github-bot

  committed
  8686109

  Copy full SHA for 8686109

Commits on Jan 25, 2021

• Merge CompilationUnit from torch._C and torch.jit (#50614)

  Show description for ac0a3cc
  

  t-vi

  authored and

  facebook-github-bot

  committed
  ac0a3cc

  Copy full SHA for ac0a3cc

Commits on Jan 9, 2021

• JIT: guard DifferentiableGraph node (#49433)

  Show description for ea087e2
  

  t-vi

  authored and

  facebook-github-bot

  committed
  ea087e2

  Copy full SHA for ea087e2

Commits on Jan 6, 2021

• Remove cpu half and dead code from multinomial (#50063)

  Show description for def8aa5
  

  t-vi

  authored and

  facebook-github-bot

  committed
  def8aa5

  Copy full SHA for def8aa5

Commits on Jan 4, 2021

• move to non-legacy magma v2 headers (#49978)

  Show description for 6e84a01
  

  t-vi

  authored and

  facebook-github-bot

  committed
  6e84a01

  Copy full SHA for 6e84a01

Commits on Nov 2, 2020

• Add RAII DetectAnomalyGuard (#47164)

  Show description for b5a1be0
  

  t-vi

  authored and

  facebook-github-bot

  committed
  b5a1be0

  Copy full SHA for b5a1be0

Commits on Oct 30, 2020

• Add anomaly mode for C++ (#46981)

  Show description for a81572c
  

  t-vi

  authored and

  facebook-github-bot

  committed
  a81572c

  Copy full SHA for a81572c

Commits on Oct 13, 2020

• Use storage.cpu() for moving storage to CPU in serialization. (#46028)

  Show description for 7b7f251
  

  t-vi

  authored and

  facebook-github-bot

  committed
  7b7f251

  Copy full SHA for 7b7f251

Commits on Oct 8, 2020

• Enable CUDA Fuser for ROCm (#45965)

  Show description for d3d8da7
  

  t-vi

  authored and

  facebook-github-bot

  committed
  d3d8da7

  Copy full SHA for d3d8da7

Commits on Oct 6, 2020

• update persons of interest (#45803)

  Show description for 7eb0a71
  

  t-vi

  authored and

  facebook-github-bot

  committed
  7eb0a71

  Copy full SHA for 7eb0a71

Commits on Oct 5, 2020

• Enable TorchBind tests on ROCm (#45426)

  Show description for 3ab88c3
  

  t-vi

  authored and

  facebook-github-bot

  committed
  3ab88c3

  Copy full SHA for 3ab88c3

Commits on Sep 29, 2020

• ROCm {emoji:2764} TensorExpr (#45506)

  Show description for 22a34bc
  

  t-vi

  authored and

  facebook-github-bot

  committed
  22a34bc

  Copy full SHA for 22a34bc

Commits on Sep 14, 2020

• Add HIP/ROCm version to collect_env.py (#44106)

  Show description for bd257a1
  

  t-vi

  authored and

  facebook-github-bot

  committed
  bd257a1

  Copy full SHA for bd257a1

Commits on Sep 8, 2020

• test_ops: skipTest only takes a single argument (#44181)

  Show description for 7c61f57
  

  t-vi

  authored and

  facebook-github-bot

  committed
  7c61f57

  Copy full SHA for 7c61f57

Commits on Sep 3, 2020

• Mark bucketize as not subject to autograd (#44102)

  Show description for 42f9897
  

  t-vi

  authored and

  facebook-github-bot

  committed
  42f9897

  Copy full SHA for 42f9897

Commits on Jul 17, 2020

• ROCm: Fix linking of custom ops in load_inline (#41257)

  Show description for 0f78e59
  

  t-vi

  authored and

  facebook-github-bot

  committed
  0f78e59

  Copy full SHA for 0f78e59

Commits on Jul 9, 2020

• Print raising warnings in Python rather than C++ if other error occurs (#41116)

  Show description for a318234
  

  t-vi

  authored and

  facebook-github-bot

  committed
  a318234

  Copy full SHA for a318234

Commits on Jul 8, 2020

• use PYTORCH_ROCM_ARCH to set GLOO_ROCM_ARCH (#40170)

  Show description for a8bc754
  

  t-vi

  authored and

  facebook-github-bot

  committed
  a8bc754

  Copy full SHA for a8bc754

Commits on Jun 22, 2020

• ROCm: Use correct device type when exporting tensors to DLPack (#40124)

  Show description for 52f3a09
  

  t-vi

  authored and

  facebook-github-bot

  committed
  52f3a09

  Copy full SHA for 52f3a09

Commits on Apr 20, 2020

• ROCm: don't warn about CUDA compute capabilities (#35949)

  Show description for dc4d888
  

  t-vi

  authored and

  facebook-github-bot

  committed
  dc4d888

  Copy full SHA for dc4d888

Commits on Apr 13, 2020

• ROCm: enable cpp_extensions.load/load_inline (#35897)

  Show description for d070c0b
  

  t-vi

  authored and

  facebook-github-bot

  committed
  d070c0b

  Copy full SHA for d070c0b

Commits on Feb 24, 2020

• catch and propagate warnings for JIT ScriptMethods (#33010)

  Show description for 481e7f2
  

  t-vi

  authored and

  facebook-github-bot

  committed
  481e7f2

  Copy full SHA for 481e7f2

Commits on Nov 15, 2019

• Add `torch.version.hip` from cmake (#29815)

  Show description for 7889e1e
  

  t-vi

  authored and

  facebook-github-bot

  committed
  7889e1e

  Copy full SHA for 7889e1e

Commits on Oct 30, 2019

• Include hierarchy information in C++ API loading error messages (#28499)

  Show description for 2526f97
  

  t-vi

  authored and

  facebook-github-bot

  committed
  2526f97

  Copy full SHA for 2526f97

Commits on Oct 29, 2019

• Remove autograd copy_ specific isFloatingPoint (#28279)

  Show description for e57a119
  

  t-vi

  authored and

  facebook-github-bot

  committed
  e57a119

  Copy full SHA for e57a119

Commits on Oct 23, 2019

• Change activation modules in C++ from using Tensor& to Tensor (#28501)

  Show description for 09ad464
  

  t-vi

  authored and

  facebook-github-bot

  committed
  09ad464

  Copy full SHA for 09ad464

Commits on Oct 15, 2019

• Use grad_out for cudnn CTC loss (#27039)

  Show description for f461184
  

  t-vi

  authored and

  facebook-github-bot

  committed
  f461184

  Copy full SHA for f461184

Commits on Oct 10, 2019

• Fix native ctc_loss gradient indexing bug for large target sizes (#27460)

  Show description for e66e00c
  

  t-vi

  authored and

  facebook-github-bot

  committed
  e66e00c

  Copy full SHA for e66e00c

Commits on Jul 31, 2019

• Fix CTC loss for zero-length targets on GPU (#23298)

  Show description for 2e40857
  

  t-vi

  authored and

  facebook-github-bot

  committed
  2e40857

  Copy full SHA for 2e40857

Commits on Jul 25, 2019

• Disable fusion of grad_sum_to_size (#23372)

  Show description for cf50249
  

  t-vi

  authored and

  facebook-github-bot

  committed
  cf50249

  Copy full SHA for cf50249

Commits on Jul 23, 2019

• Add Module::replace_module to C++ api (#22546)

  Show description for 0dabaad
  

  t-vi

  authored and

  facebook-github-bot

  committed
  0dabaad

  Copy full SHA for 0dabaad

Commits on Jul 22, 2019

• hipify: do not overwrite files that stay the same (#23112)

  Show description for ff8cb9f
  

  t-vi

  authored and

  facebook-github-bot

  committed
  ff8cb9f

  Copy full SHA for ff8cb9f

Pagination

Previous

Next