strisunshinewentingwang/pytorch

Pull Requests Commits

[ONNX] Update ORT version (#41372)

neginraoof committed 4 years ago

319b20b7

[ONNX] Export embedding_bag (#41234)

neginraoof committed 4 years ago

346c69a6

Profiler: Do not record zero duration kernel events (#41540)

mwootton committed 4 years ago

7eb71b4b

fix division by low precision scalar (#41446)

ngimel committed 4 years ago

324c18fc

[JIT] Teach IRPrinter and IRParser to handle 'requires_grad' and 'device' as a part of type info. (#41507)

Mikhail Zolotukhin committed 4 years ago

5d704652

Adding missing setting `state_.ptr()` and `hook_.ptr()` to `nullptr`. (#41537)

sinannasir committed 4 years ago

241bc648

Initial implementation of quantile operator (#39417)

heitorschueroff committed 4 years ago

c7798ddf

Add `torch.atleast_{1d/2d/3d}` (#41317)

kshitij12345 committed 4 years ago

71fdf748

Add reference documentation for torch/library.h (#41470)

ezyang committed 4 years ago

840ad94e

rewrite C++ __torch_function__ handling to work with TensorList operands (#41575)

ngoldbaum committed 4 years ago

1e230a5c

Assert valid inner type for OptionalType creation (#41509)

wconstab committed 4 years ago

cb9029df

enable jit profiling tests on macos (#41550)

Krovatkin committed 4 years ago

e3e58e20

During inbatch broadcast, move Tile op after Fused8BitRowwiseQuantizedToFloat if applicable (#41464)

Yinghai Lu committed 4 years ago

eb3bf96f

Run NO_AVX jobs on CPU (#41565)

malfet committed 4 years ago

5376785a

[JIT] make fastrnns runnable on cpu (#41483)

Elias Ellison committed 4 years ago

728fd37d

Revert D22552377: [pytorch][PR] Reland split unsafe version

albanD committed 4 years ago

b1d4e33c

Create lazy_dyndeps to avoid caffe2 import costs. (#41343)

c00w committed 4 years ago

415ff0bc

Use c10::cuda:: primitives rather than make CUDA runtime calls directly (#41405)

maokaiyu committed 4 years ago

9ed82574

Makes the use of the term "module" consistent through the serialization note (#41563)

Mike Ruberry committed 4 years ago

a0e58996

Fix RocM resource class allocation (#41553)

malfet committed 4 years ago

454cd3ea

Add tests to logical operation in BinaryOpsKernel.cpp (#41515)

scintiller committed 4 years ago

e324ea85

Notes for lcm and gcd, formatting doc fixes (#41526)

Mike Ruberry committed 4 years ago

f49d97a8

Revert D22519869: [pytorch][PR] RandomSampler generates samples one at a time when replacement=True

mrshenli committed 4 years ago

86590f22

[RocM] Switch to rocm-3.5.1 image (#41273)

malfet committed 4 years ago

ba6b2354

RandomSampler generates samples one at a time when replacement=True (#40026)

ydaiming committed 4 years ago

09647e12

[Gloo] alltoall to ProcessGroupGloo (#41424)

jiayisuse committed 4 years ago

6f5f455c

Remove unnecessary test in rpc_test.py (#41218)

rohan-varma committed 4 years ago

1ac46924

Fix flaky test_udf_remote_message_delay_timeout_to_self (#41217)

rohan-varma committed 4 years ago

b5e32528

Split ASAN and ROCM tests into test1 and test2 (#41520)

malfet committed 4 years ago

94e4248d

[Gloo] Tests for Gloo Async Work Wait-level Timeouts (#41265)

osalpekar committed 4 years ago

81e96490

Newer Older