strisunshinewentingwang/pytorch

Pull Requests Commits

[ONNX] Update merge rules and persons of interest (#82673)

BowenBao committed 2 years ago

26d50ff1

add saketh-are to torch.nn codeowners (#79431)

saketh-are committed 2 years ago

a1af0d1b

Allow user to assert no mask contiguous check is necessary (#82533)

mikekgfb committed 2 years ago

82f558fe

Implement and extend `mul(sparse, sparse)` to work with 0-dim arguments on either side. (#82717)

nikitaved committed 2 years ago

3ab54b97

Remove unnecessary `import warnings` (#82760)

kurtamohler committed 2 years ago

8b4fee59

[primTorch] Added unbind OpInfo and ref (#81776)

Fabio Rocha committed 2 years ago

ff753cbc

Add torch.ops.nvprims namespace for nvFuser-specific prims (#82155)

IvanYashchuk committed 2 years ago

ec67c6ab

Fix objcopy version detection (#82774)

malfet committed 2 years ago

7c298b82

Fix NCCL detection by Gloo (#82773)

malfet committed 2 years ago

83086b7f

Fix device behavior for masked_fill (#82737)

ngimel committed 2 years ago

112ec24f

Fix Windows builds with _DEBUG flag

malfet committed 2 years ago

Verified bbe8d019

Add DLPack support for XPU backend by mapping to kDLOneAPI in DLPack. (#81021)

chengjunlu committed 2 years ago

8be85302

[xla hash update] update the pinned xla hash (#82805)

pytorchmergebot committed 2 years ago

bcc8f592

Add a line about module owners to labels mapping in CODEOWNERS (#82748)

janeyx99 committed 2 years ago

3687792c

[torchgen] Generate wrapper functions under custom namespaces (#81744)

larryliu0820 committed 2 years ago

406ce692

[_shard] only check shard metadata for copy_ (#82655)

wanchaol committed 2 years ago

cda8635a

[torchdynamo hash update] update the pinned torchdynamo hash (#82458)

pytorchmergebot committed 2 years ago

0a919e8b

Store compilation os version on a per model basis (#82661)

jmdetloff committed 2 years ago

7d5db630

Register Triton scaled dot production in aten (#82509) (#82758)

zrphercule committed 2 years ago

dd838cee

Revert "Revert "Added zero.symint and modified aten::trapz to use symbolic ints (#82054)"" (#82779)

Chillee committed 2 years ago

1164c83c

Made another tweak to inductor heuristic + made it enableable from an arg (#82772)

Chillee committed 2 years ago

95ed516d

Revert "Bug fix in ExtraCUDACopy and remove unstable lints for release (#82693)"

pytorchmergebot committed 2 years ago

d852dce7

Add nondeterministic alert to CUDA cumsum (#75693)

kurtamohler committed 2 years ago

c3799159

Enable `dim=None` for `torch.std` (#81845)

kurtamohler committed 2 years ago

eb0e30e0

Add ref for meshgrid (#82284)

soulitzer committed 2 years ago

ed6d2b56

Fix embedding quantization issue when memory format is not `contiguous ` (#82605)

sijiac committed 2 years ago

26d4e531

Properly compare floats in fully-connected-sparse-operator-tester (#82768)

kit1980 committed 2 years ago

9887d51c

Revert "Added zero.symint and modified aten::trapz to use symbolic ints (#82054)"

pytorchmergebot committed 2 years ago

0f52794c

[Vulkan] Implement packed contexts (#82730)

manuelcandales committed 2 years ago

9ec8d64d

[ONNX] Inline prim::PythonOp for Autograd Function Export (#74765)

shubhambhokare1 committed 2 years ago

95d87385

Newer Older