Commits

Branch selector

 master

User selector

t-vi

Datepicker

All time

Commit History

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

Commits on Jun 26, 2019

• export deleteFunction in torch/csrc/autograd/function.h (#22236)

  Show description for c1fc2f2
  

  t-vi

  authored and

  facebook-github-bot

  committed
  c1fc2f2

  Copy full SHA for c1fc2f2

Commits on Jun 25, 2019

• Don't end on inplace operators in einsum (#22111)

  Show description for f5df0c9
  

  t-vi

  authored and

  facebook-github-bot

  committed
  f5df0c9

  Copy full SHA for f5df0c9

Commits on Jun 6, 2019

• pack_padded_sequence: Check for empty (zero-element) tensors (#21461)

  Show description for 3feb40d
  

  t-vi

  authored and

  facebook-github-bot

  committed
  3feb40d

  Copy full SHA for 3feb40d

Commits on May 29, 2019

• ctc_loss: fix backward when 2d target tensor is larger than max_target_length (#20971)

  Show description for aa42742
  

  t-vi

  authored and

  facebook-github-bot

  committed
  aa42742

  Copy full SHA for aa42742

• Allow nondet_tol for nondeterminism in gradcheck and gradgradcheck (#20980)

  Show description for d23d04f
  

  t-vi

  authored and

  facebook-github-bot

  committed
  d23d04f

  Copy full SHA for d23d04f

Commits on May 24, 2019

• JIT: Eliminate SumToSize by using Optional Lists (#18697)

  Show description for 17941f9
  

  t-vi

  authored and

  facebook-github-bot

  committed
  17941f9

  Copy full SHA for 17941f9

Commits on May 8, 2019

• tweak scripts/build_android.sh for ABI and header install (#20152)

  Show description for 3ac4d92
  

  t-vi

  authored and

  facebook-github-bot

  committed
  3ac4d92

  Copy full SHA for 3ac4d92

Commits on May 6, 2019

• Specialize Optional[T] to T (or subtype for Tensor) or None when executing graph (#18407)

  Show description for 5c9ab6f
  

  t-vi

  authored and

  facebook-github-bot

  committed
  5c9ab6f

  Copy full SHA for 5c9ab6f

Commits on Apr 25, 2019

• Fall back to asking nvcc for detecting cuda version if no *cudaart* is found (#19741)

  Show description for 556c8a3
  

  t-vi

  authored and

  facebook-github-bot

  committed
  556c8a3

  Copy full SHA for 556c8a3

Commits on Apr 19, 2019

• pow scalar exponent / base autodiff, fusion (#19324)

  Show description for b9291f5
  

  t-vi

  authored and

  facebook-github-bot

  committed
  b9291f5

  Copy full SHA for b9291f5

Commits on Apr 8, 2019

• Handle None indexing in TorchScript (#18615)

  Show description for 026a9c6
  

  t-vi

  authored and

  facebook-github-bot

  committed
  026a9c6

  Copy full SHA for 026a9c6

Commits on Mar 29, 2019

• Update NNPACK to current master (#18580)

  Show description for b1a0233
  

  t-vi

  authored and

  facebook-github-bot

  committed
  b1a0233

  Copy full SHA for b1a0233

Commits on Mar 28, 2019

• Use __ldg for CUDA kernels in fuser (#18540)

  Show description for 9696f06
  

  t-vi

  authored and

  facebook-github-bot

  committed
  9696f06

  Copy full SHA for 9696f06

Commits on Mar 25, 2019

• Revert "Specialize optional tensor inputs to graphs in the JIT (#18360)" (#18411)

  Show description for 9e176fe
  

  t-vi

  authored and

  facebook-github-bot

  committed
  9e176fe

  Copy full SHA for 9e176fe

Commits on Mar 24, 2019

• Specialize optional tensor inputs to graphs in the JIT (#18360)

  Show description for 7cc7ed1
  

  t-vi

  authored and

  facebook-github-bot

  committed
  7cc7ed1

  Copy full SHA for 7cc7ed1

Commits on Mar 22, 2019

• Fix error reporting in NVRTC use of the fuser (#18327)

  Show description for 2aac180
  

  t-vi

  authored and

  facebook-github-bot

  committed
  2aac180

  Copy full SHA for 2aac180

Commits on Mar 21, 2019

• move median to ATen (#17637)

  Show description for 8356ffa
  

  t-vi

  authored and

  facebook-github-bot

  committed
  8356ffa

  Copy full SHA for 8356ffa

• Two amendments for the shape analysis (#18271)

  Show description for 0122121
  

  t-vi

  authored and

  facebook-github-bot

  committed
  0122121

  Copy full SHA for 0122121

Pagination

Previous

Next