strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D25239967: [fx] Add matrix multiplication fusion pass

ngimel committed 3 years ago

ad7d208b

[PyTorch] Reapply D25546409: Use .sizes() isntead of .size() in cat_serial_kernel_impl (#49762)

swolchok committed 3 years ago

282552dd

Fix for possible RNG offset calculation bug in cuda vectorized dropout with VEC=2 (#50110)

mcarilli committed 3 years ago

57d489e4

[reland][quant][graphmode][fx] Standalone module support {input/output}_quantized_idxs (#49754) (#50058)

jerryzh168 committed 3 years ago

f6f0fde8

Fix mypy typing check for test_dataset (#50108)

ejguan committed 3 years ago

05358332

Remove cpu half and dead code from multinomial (#50063)

t-vi committed 3 years ago

def8aa54

[fx] Add matrix multiplication fusion pass (#50120)

Meghan Lele committed 3 years ago

9b7f3fa1

Clean up type annotations in caffe2/torch/nn/modules (#49957)

r-barnes committed 3 years ago

d80d38cf

[PyTorch] Reapply D25545777: Use .sizes() instead of .size() in _cat_out_cpu (#49761)

swolchok committed 3 years ago

75028f28

[PyTorch] Reapply D25544731: Avoid extra Tensor refcounting in _cat_out_cpu (#49760)

swolchok committed 3 years ago

574a15b6

Fix doc for vmap levels (#50099)

ejguan committed 3 years ago

5f875965

Kill AT_SKIP_BFLOAT16_IF_NOT_ROCM (#48810)

zasdfgbnm committed 3 years ago

70734f12

Support out argument in torch.fft ops (#49335)

peterbell10 committed 3 years ago

26391143

torch.flip and torch.flip{lr, ud}: Half support for CPU and BFloat16 support for CPU & CUDA (#49895)

kshitij12345 committed 3 years ago

5d93e2b8

fix fork formatting (#49436)

Elias Ellison committed 3 years ago

d1c375f0

Revert D25746115: [pytorch][PR] Improve documentation and warning message for creation of a tensor with from_numpy()

Mike Ruberry committed 3 years ago

7fe25af5

[PyTorch Mobile] Mark xnnpack operators selective

iseeyuan committed 3 years ago

dcc83868

Make stft (temporarily) warn (#50102)

Mike Ruberry committed 3 years ago

5e1c8f24

Improve documentation and warning message for creation of a tensor with from_numpy() (#49516)

leonvol committed 3 years ago

4a6c178f

Revert D25757721: [pytorch][PR] Run mypy on more test files

Mike Ruberry committed 3 years ago

9529ae37

[docs] add docstring in torch.cuda.get_device_properties (#49792)

Jeff Yang committed 3 years ago

d1a56fcd

[JIT] Add `__prepare_scriptable__` duck typing to allow replacing nn.modules with scriptable preparations (#45645) (#49242)

Meghan Lele committed 3 years ago

abe1fa49

[pytorch][codegen] migrate gen_variable_type to new data model (#49735)

ljk53 committed 3 years ago

e71a13e8

[PyTorch] Avoid heap allocations in inferUnsqueezeGeometry (#49497)

swolchok committed 3 years ago

a272a7ee

Enable distribution validation if __debug__ (#48743)

fritzo committed 3 years ago

093aca08

Revert D25757691: [pytorch][PR] Run mypy over test/test_utils.py

Rong Rong (AI Infra) committed 3 years ago

e3c56ddd

Update MultiHeadAttention docstring (#49950)

vfdev-5 committed 3 years ago

e442ac1e

Drop unused imports from caffe2/python (#49980)

r-barnes committed 3 years ago

9945fd72

[caffe2][a10] Move down pragma pop to properly suppress warning 4522 (#49233)

jdonald committed 3 years ago

eee849be

Fix store based barrier to only use 'add'. (#49930)

pritamdamania committed 3 years ago

16e5af41

Newer Older