strisunshinewentingwang/pytorch

Pull Requests Commits

Latest linalg_lu revert caused bc to fail, add exceptions (#77112)

janeyx99 committed 2 years ago

0272aa1f

Fix docker builds, cleanup cuda 115 (#77086)

atalman committed 2 years ago

8a6856ae

make has_bundled_input work for flatbuffer (#76854)

qihqi committed 2 years ago

41ff6f8c

Skip lu_solve meta tests (#77110)

malfet committed 2 years ago

b4fa1e88

disabling squeeze/unsqueeze; disabling BN/BN_BWD for perf concern (#77017)

jjsjann123 committed 2 years ago

489818e7

[distributions] Apply weaker test for `torch.distributions.wishart.Wishart` to match the level in other libraries. (#76525)

nonconvexopt committed 2 years ago

84fe6021

[sr][pyper] add fusion broadcast_concat_batch_matmul_batch_gather (#76839)

ajyu committed 2 years ago

eb0bc5b1

shard `linux-xenial-cuda11.3-py3.7-gcc7 / test (default` from 2 to 4 (#77083)

clee2000 committed 2 years ago

a28e4366

[JIT SSA] Added testing for the Cat Op in LazyTensor

Gamrix committed 2 years ago

26e2936e

[RecordFunction] Store a c10::variant of name and schema rather then both. (#76017)

Taylor Robie committed 2 years ago

8355c660

Revert "Move Tensor.grad back into C++"

pytorchmergebot committed 2 years ago

0c7c5097

Revert "[LT] Store OpKind for each IR subclass in a static field"

pytorchmergebot committed 2 years ago

7eaf4780

[composite compliance] backward: value selecting reduction ops (#76731)

kshitij12345 committed 2 years ago

da565c07

[fx] fix merge_matmul tests making invalid torch.split calls

peterbell10 committed 2 years ago

b9e919fe

Revert "stft: remove non-center overload and python functional wrapper"

pytorchmergebot committed 2 years ago

2c5bf125

Faster `index_select` for sparse COO tensors on CPU. (#72710)

nikitaved committed 2 years ago

ce3857e7

Move Tensor.grad back into C++

peterbell10 committed 2 years ago

3e4bff72

Revert "Add linalg.lu_solve"

pytorchmergebot committed 2 years ago

4ebc4890

Revert "Update torch.lu_unpack docs"

pytorchmergebot committed 2 years ago

4ceac494

Revert "Deprecate torch.lu"

pytorchmergebot committed 2 years ago

1467e0dd

Revert "Deprecate torch.lu_solve"

pytorchmergebot committed 2 years ago

b042cc7f

Revert "Move `contiguous_strides` to `c10/util/strides.h`."

pytorchmergebot committed 2 years ago

13a84717

[PyTorch] Don't enter MHA fast path when bias & query dtypes don't match

swolchok committed 2 years ago

e5915a22

Revert "Revert "[WIP] customize the C++ class for valueT"" (#77003)

Krovatkin committed 2 years ago

daf8c48a

LT tutorial [WIP] (#76392)

Krovatkin committed 2 years ago

a6341d2c

[LT] Move device lock in LazyGraphExecutor to a later place

desertfire committed 2 years ago

36150c63

[Bootcamp]Add option for flatbuffer loader to copy memory to individual tensors (#76986)

fuqianz committed 2 years ago

bd573389

[PyTorch] Replace IValue::intrusive_ptr with IValue::isIntrusivePtr()

swolchok committed 2 years ago

0b0611c2

[PyTorch] Add corner case tests for IValue::use_count()

swolchok committed 2 years ago

c3e5c120

Torchvision patch (#77001)

jjsjann123 committed 2 years ago

b4f3f9c6

Newer Older