strisunshinewentingwang/pytorch

Pull Requests Commits

[DataPipe] Fixes an issue where TarArchiveReader closes stream when read into a buffer

NivekT committed 2 years ago

756331f9

[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

CMake: Limit python include directories to only python libraries (#65654)

peterbell10 committed 2 years ago

971c57f1

[Static Runtime] concat_add_mul_replacenan_clip retains axis arg (#65741)

Mike Iovine committed 2 years ago

5f7ab7be

[PyTorch Edge] Reduce the cost of computing isIncludedInAlias() (#65735)

dhruvbird committed 2 years ago

f63150fd

deprecate device getter from `torch.testing` namespace (#63844)

pmeier committed 2 years ago

aebde1bc

move kernel launch checks from `torch.testing` to `torch.testing._internal.check_kernel_launches` (#60862)

pmeier committed 2 years ago

07d5d7b5

Revert D31206837: [pytorch][PR] `*_solve` methods: implements forward AD

Mike Ruberry committed 2 years ago

0a0564a3

Older