strisunshinewentingwang/pytorch

Pull Requests Commits

[NNC] Skip buildShapeExpressions if ConstantChunk input shapes are unknown (#82698)

davidberard98 committed 2 years ago

b721ea9b

[NestedTensor]Remove tensor buffer replace with Storage type (#82757)

drisspg committed 2 years ago

2f9d046d

Test that multi_tensor optimizer state buffers match with single_tensor state buffers (#81894)

mikaylagawarecki committed 2 years ago

b1922e03

Revert "Re-land sym_numel (#82374) (#82726) (#82731)"

pytorchmergebot committed 2 years ago

78bd95b1

[functorch] Add a bunch of low hanging fruit linalg batch rules (#82177)

samdow committed 2 years ago

ae399d00

Refactor quant levels visualization (#82790)

asl3 committed 2 years ago

34103a30

Fix a segfault in `new_empty_strided` (#82422)

Nikita Karetnikov committed 2 years ago

f6749e76

Skip tests using uninitialized data (#82728)

Nikita Karetnikov committed 2 years ago

c1b9c953

[Vulkan] Implement arithmetic ops where one of the arguments is a tensor wrapping a scalar. (#81726)

manuelcandales committed 2 years ago

0d22d684

making sure we don't accept pointers in int64_t c-tor (#82743)

Krovatkin committed 2 years ago

9958cbee

Re-land sym_numel (#82374) (#82726) (#82731)

Krovatkin committed 2 years ago

c90e00cf

Enable `dim=None` for `torch.var` (#82765)

kurtamohler committed 2 years ago

5ca9b2b6

Adjust schemas of a couple prims (#82821)

eellison committed 2 years ago

e3dd4242

add integer divison for symints (#82791)

Krovatkin committed 2 years ago

8b20e479

Update NCCL to v2.13.4-1 (#82775)

malfet committed 2 years ago

c08092fd

dump ios test simulator logs on a failure (#82822)

Krovatkin committed 2 years ago

61b1301e

[vulkan] Add buffer to texture and texture to buffer copies (#82799)

SS-JIA committed 2 years ago

8e33396c

[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

Newer Older