strisunshinewentingwang/pytorch

Pull Requests Commits

Fix typing errors in torch.distributed.*, close issue #42967. (#47534)

xuzhao9 committed 3 years ago

49f0e5df

Fix typing errors in torch.distributed.nn.* directory. (#47533)

xuzhao9 committed 3 years ago

7f66fa62

Fix typing errors in torch.distributed.distributed_c10d.* (#47532)

xuzhao9 committed 3 years ago

915050ed

Fix type annotation errors in torch.distributed.* directory (#47531)

xuzhao9 committed 3 years ago

49eb82a7

[pytorch][te] Do not merge Tensor[] variant of aten::where into fusion group (#48063)

bertmaher committed 3 years ago

af37f8f8

[TorchScript] Support user defined classes as constants (#5062)

bwasti committed 3 years ago

43a9d6fb

[JIT] Optimize FunctionSchema::checkArg for the Tensor case. (#48034)

Mikhail Zolotukhin committed 3 years ago

3611d26a

Revert D24714803: make duplicate def() calls an error in the dispatcher. Updating all fb operators to use the new dispatcher registration API

bdhirsh committed 3 years ago

7b2c78f1

[caffe][memonger] Extend operator schema check to dag memonger (#48021)

singlaiiit committed 3 years ago

549ef1d6

fix node manipulation in partition class (#48016)

scottxu0730 committed 3 years ago

fa0acb73

make duplicate def() calls an error in the dispatcher. Updating all fb operators to use the new dispatcher registration API (#47322)

bdhirsh committed 3 years ago

824f7106

migrate export_caffe2_op_to_c10.h macros to the new dispatcher registration API (#47321)

bdhirsh committed 3 years ago

cba26e40

rename macro. TORCH_LIBRARY_FRAGMENT_THIS_API_IS_FOR_PER_OP_REGISTRATION_ONLY to TORCH_LIBRARY_FRAGMENT (#47320)

bdhirsh committed 3 years ago

93d98373

update legacy dispatcher registration API tests to avoid duplicate def() calls (#47319)

bdhirsh committed 3 years ago

95b9c206

remove ops in the __caffe2 namespace (#47318)

bdhirsh committed 3 years ago

6ec2a89e

Make sure valid ParameterList/Dict don't warn on creation (#47772)

albanD committed 3 years ago

233192be

Test TORCH_LIBRARY in CUDA extension (#47524)

zasdfgbnm committed 3 years ago

b12d645c

add type annotations to multiprocessing module (#47756)

guilhermeleobas committed 3 years ago

cf92b0f3

Fix docstring typo (#47545)

OverLordGoldDragon committed 3 years ago

1e0ace7f

[caffe2] plan_executor_test: add test case for should_stop loops (#47613)

d4l3k committed 3 years ago

825ee7e7

Port math kernel for layer_norm from pytorch/xla. (#47882)

Ailing Zhang committed 3 years ago

550f26c6

Set proper output differentiability for unique function (#47930)

albanD committed 3 years ago

95ea778a

torch.Assert: make it torch.jit.script'able (#47399) (#47973)

vkuzo committed 3 years ago

dea23378

rename torch.Assert to torch._assert (#47763) (#47972)

vkuzo committed 3 years ago

ee995d33

Skip dummy node creation for autograd engine when there is a single input and place on correct queue (#47592)

soulitzer committed 3 years ago

d20483a9

[NNC] Support vectorization of reductions (#47924)

nickgg committed 3 years ago

957e45a9

[CI] Fix additional CI jobs not launched when PR is created from fork repo (#47969)

Rong Rong committed 3 years ago

9aaf7fb3

Fix documentation to point to torch.overrides instead of _overrides. (#47842)

hameerabbasi committed 3 years ago

3a2aad93

update windows build guide (#47840)

mszhanyi committed 3 years ago

f9552e6d

[cmake] clean up cmake/Utils.cmake (#47923)

Rong Rong committed 3 years ago

147a48fb

Newer Older