strisunshinewentingwang/pytorch

Pull Requests Commits

quantize bias of the quantization parameters (#48749)

Hector Yuen committed 3 years ago

b726a1bb

Remove output used only by sizes (#448) (#47665)

jjsjann123 committed 3 years ago

dabc286a

Add nondeterministic alert to index_copy, median CUDA and kthvalue CUDA (#46942)

kurtamohler committed 3 years ago

2cb92041

Add scary comment in cpp_custom_type_hack.h (#48737)

James Reed committed 3 years ago

c2ad3c4e

[Pytorch][Annotation] Update inlined callstack with module instance info (#47416)

cccclai committed 3 years ago

416dc683

[numpy] Add `torch.moveaxis` (#48581)

kshitij12345 committed 3 years ago

5c9cef9a

[ONNX] Cast Gather index to Long if needed (#47653)

jiafatom committed 3 years ago

befab0d9

Enable TCPStore on Windows (#47749)

gunandrose4u committed 3 years ago

92f37614

Header cleanup (#48728)

ezyang committed 3 years ago

93973ee6

Fix code review from #48659 and #48116 (#48731)

ezyang committed 3 years ago

f9a0abfc

Generalize some TensorIterator consumers to take TensorIteratorBase (#48727)

ezyang committed 3 years ago

d6f9e856

Simplify CachingAllocator. (#48752)

kimishpatel committed 3 years ago

c01e5b88

reenabling MPI test (#48725)

Rong Rong committed 3 years ago

ef50c94e

Replace constexpr with CONSTEXPR_EXCEPT_WIN_CUDA (#48717)

datumbox committed 3 years ago

0484b048

Add support for CorrCholeskyTransform (#48041)

neerajprad committed 3 years ago

5489a98c

Add broadcast_shapes() function and use it in MultivariateNormal (#43935)

fritzo committed 3 years ago

313e77fc

Revert D24874754: [pytorch][PR] Add test for empty tensors for batch matmuls

Mike Ruberry committed 3 years ago

c7746adb

Optimize torch zeros (#45636)

Edvard Ghazaryan committed 3 years ago

79b9c034

Fix unintended error when worker force kill happens #43455 (#43462)

ssnl committed 3 years ago

1112773c

Replace kernel resource strings with real .cu source files (#48283)

tlemo committed 3 years ago

85c1e8ac

Add test for empty tensors for batch matmuls (#47700)

zasdfgbnm committed 3 years ago

5f105e2a

[qunat][graphmode][fx] Standalone module takes float as input and output (#48671)

jerryzh168 committed 3 years ago

ea573ea9

Disable autocast cache for tensor views as fix for #48049 (#48696)

pbialecki committed 3 years ago

22c3ae8b

Refactored OpInfo testing to support custom SampleInputs, added addmm to op_db to test (#48627)

Lilyjjo committed 3 years ago

0e4f9a78

Support for OpInfo-based testing for operators in JIT (#47696)

Lilyjjo committed 3 years ago

90faf431

Refactored assertAutodiff test to have better error message (#48567)

Lilyjjo committed 3 years ago

9c35a680

Refactor existing JIT testing utils to enable new OpInfo test suite to reuse existing logic (#47695)

Lilyjjo committed 3 years ago

c465602d

[NNC] Add cpu fusion gflag (#48682)

Elias Ellison committed 3 years ago

11954039

only enable cudnn persistent RNN when batchsize % 8 == 0 (#48070)

xwang233 committed 3 years ago

0d39bd47

Use fastAtomicAdd in GPU upsampling trilinear (#48675)

xwang233 committed 3 years ago

a1daf1e6

Newer Older