strisunshinewentingwang/pytorch

Pull Requests Commits

[DataPipe] Adding interface for MapDataPipes (#69648)

NivekT committed 2 years ago

6713241c

[DataPipe] Adding output types to DataPipe interface file (#69647)

NivekT committed 2 years ago

667f0ff1

[PyTorch] Update SmallVector from LLVM (#69110)

swolchok committed 2 years ago

8d37cb65

[PyTorch] Simple refcount bump fixes in standardizeVectorForUnion & callees (#66695)

swolchok committed 2 years ago

7c360be3

[shard] fix with_comms wrapper (#69493)

wanchaol committed 2 years ago

221101e3

[JIT] simplify logic for merging types during profiling (#69096)

davidberard98 committed 2 years ago

e3e7a7c9

Make vec512 bfloat16 map function clang-Wall clean (#69707)

malfet committed 2 years ago

6a2f3b63

Cleanup ProcessGroup.cpp (#69706)

malfet committed 2 years ago

421bbc3b

Make blend operations clang-Wall clean (#69705)

malfet committed 2 years ago

91ced444

fix CompositeImplicitAutograd ops improperly labeled (#69169)

bdhirsh committed 2 years ago

8a113509

use irange for loops 2 (#66746)

r-barnes committed 2 years ago

22feaf41

[Jit] Fix schema of aten::split int[] version (#69745)

Hao Lu committed 2 years ago

8f3474c5

Remove THTensor (#69040)

peterbell10 committed 2 years ago

b2191049

[tensorexpr] Fix the buf size of discontiguous tensors (#69657)

huiguoo committed 2 years ago

2a1e3fde

Add native masked_softmax (#69268)

zrphercule2 committed 2 years ago

c71109cf

[SR] Wrap check_for_memory_leak with DCHECK (#69588)

Hao Lu committed 2 years ago

e4d4e6ec

Make c10d tests -Werror clean (#69703)

malfet committed 2 years ago

41a163f1

Do not use `std::labs` (#69704)

malfet committed 2 years ago

8c0687d6

[ONNX] shapeValueMap: Represent symbolic shape as value (#68203) (#69545)

BowenBao committed 2 years ago

c88774e4

Back out "[wip][quant][graphmode] produce reference pattern for binary ops and then rewrite to quantized op" (#69713)

Ha-nyung Chung committed 2 years ago

5b9f78ea

[nnc][quant] Fix quantized concat (#69596)

IvanKobzarev committed 2 years ago

75c9e1a9

Remove WindowsTorchApiMacro.h in favor of Export.h (#69585)

peterbell10 committed 2 years ago

2b0b1916

[SR] assignStorageToManagedTensors returns a vector (#69568)

Mike Iovine committed 2 years ago

e459edd1

[Static Runtime] [Code Cleanup] Encapsulate function objects within ProcessedFunction (#69595)

d1jang committed 2 years ago

90340004

Introduce isTensorSubclassLike; add special cases to backwards formulas (#69534)

zou3519 committed 2 years ago

1f2fabf7

[pytorch][PR] Add ability for a mobile::Module to save as flatbuffer (#69306)

qihqi committed 2 years ago

b0fbee86

.github: Add fbsync to push triggers (#69718)

seemethere committed 2 years ago

6e496f57

[SR] Update memory planner docs (#69559)

Mike Iovine committed 2 years ago

58b53a0f

Extend explanation of `torch.cholesky_inverse` to consider batched inputs. (#69069)

nonconvexopt committed 2 years ago

c886b17e

Upgrade oneDNN to v2.3.3 and package oneDNN Graph API together (#63748)

chunyuan-w committed 2 years ago

b93bc54f

Newer Older