strisunshinewentingwang/pytorch

Pull Requests Commits

[JIT] Make output of prim::TupleConstruct alias only with its inputs (#64879)

d1jang committed 2 years ago

cd458fe0

Skip failing tests when LAPACK and MAGMA are not available (#64930)

IvanYashchuk committed 2 years ago

dd354117

Remove "SciPioneer" from PT Distributed code owners (#65862)

Yi Wang committed 2 years ago

2c29ec2a

Revert D31005792: [NCCL] Init dummy NCCL comms in constructor

Mike Ruberry committed 2 years ago

91f8755b

Migrate THCIntegerDivider.cuh to ATen (#65745)

ngimel committed 2 years ago

5349ea92

(torchelastic) make --max_restarts explicit in the quickstart and runner docs (#65838)

Kiuk Chung committed 2 years ago

3900509b

[quant] Add imports to the torch/ao/quantization/__init__.py (#64911)

Zafar Takhirov committed 2 years ago

c7ef620a

Avoid saving self for`softmax` and `log_softmax` (#65242)

soulitzer committed 2 years ago

fb412bdd

fix typo in _sharded_tensor (#65511)

crcrpar committed 2 years ago

768cfaa8

Make JIT Aliasing Test Less Brittle (#65493)

Jiewen Tan committed 2 years ago

9f97c66a

Decouple forward AD checks from backward AD in OpInfo tests and gradcheck (#65040)

soulitzer committed 2 years ago

91611fe1

Stop Win+CUDA-10.2 builds (#65649)

malfet committed 2 years ago

5950240b

[NCCL] Init dummy NCCL comms in constructor (#65173)

rohan-varma committed 2 years ago

2b22a5dd

Remove THCDeviceTensor (#65744)

ngimel committed 2 years ago

ad85b582

slow_conv2d_forward: avoid calling dispatcher in parallel region (#65724)

peterbell10 committed 2 years ago

20374c99

Update Module docstring for Python 3 (#65748)

JackKelly committed 2 years ago

7191dd26

ns for fx: add basic testing on cuda (#65593)

vkuzo committed 2 years ago

8bf0ba54

Migrate THCScanUtils to ATen (#65743)

ngimel committed 2 years ago

0dd1b74a

[PyTorch Edge] Conditionally trim dispatch key set to save heap memory at runtime (#65732)

dhruvbird committed 2 years ago

a84feeea

.github: Bump linux gpu max limit to 100 (#65831)

seemethere committed 2 years ago

7b5d676f

[Static Runtime] Simplify out variant overload implementations (#65384)

Mike Iovine committed 2 years ago

c975ca43

.github: Remove confusing on_pull_request variable (#65731)

seemethere committed 2 years ago

2f712c45

common_utils.py: Add ASAN as a platform for which you can disable tests (#65791)

janeyx99 committed 2 years ago

6c2f235d

type annotate operator_support (#65136)

kflu committed 2 years ago

911d01c1

[ROCm] Changes not to rely on CUDA_VERSION or HIP_VERSION (#65610)

pruthvistony committed 2 years ago

085e2f7b

Revert D31193205: [pytorch][PR] CMake: Limit python include directories to only python libraries

suo committed 2 years ago

9b40eaaa

LLVM-12 fix for tensor_new.cpp (#65785)

r-barnes committed 2 years ago

2670cacf

don't check 0 elements for cat symbolic diff (#65751)

ngimel committed 2 years ago

09eb3e66

Migrate THCThrustAllocator to ATen (#65492)

peterbell10 committed 2 years ago

1d681c1a

saketh-are committed 2 years ago

40ed24ec

Newer Older