strisunshinewentingwang/pytorch

Pull Requests Commits

Skip `test_minmax_illegal_dtype` on XLA (#39693)

malfet committed 4 years ago

1790d358

Fix ONNX export of RNNs with no bias (#36894)

yaeldMS committed 4 years ago

0251ba61

some refactor on register_distributed_ops (#38657)

wanchaol committed 4 years ago

9f719973

[jit] register distributed backward (#38494)

wanchaol committed 4 years ago

f32c9eb5

[dist_autograd] expose distributed backward C++ API (#38656)

wanchaol committed 4 years ago

d4939184

Enable RRef timeout for tensorpipe (#39531)

rohan-varma committed 4 years ago

e033db04

Migrate AT_DISPATCH_FLOATING_AND_COMPLEX_TYPES to c10::complex (#39296)

zasdfgbnm committed 4 years ago

afb2d27b

update convert_sync_batchnorm docs (#39646)

ssnl committed 4 years ago

d1cdf1fd

Migrate `diag` and `trace` from TH to ATen (CUDA) (#36876)

kshitij12345 committed 4 years ago

1f7557d1

Skip unit tests relying on MKL if compiled without it (#39672)

malfet committed 4 years ago

64192ca3

Remove tuple from reduction (#39433)

zasdfgbnm committed 4 years ago

8004d359

[quant][graphmode] Preserve numerics in debug option for clamp ops (#39219)

jerryzh168 committed 4 years ago

9551fb22

Cleanup unused args in max_unpooling3d (#39664)

Ailing Zhang committed 4 years ago

dd5aa1fb

setup: Add long description to wheel packages (#39676)

seemethere committed 4 years ago

b7b74335

.circleci: Fold postnightly workfow into nightly (#39669)

seemethere committed 4 years ago

84d8d683

[TensorPipe Agent] Documentation fixes and nits (#39467)

osalpekar committed 4 years ago

0147216a

Add undefined tensor gradient support to all backward functions (#39400)

kurtamohler committed 4 years ago

bba30d1b

.circleci: Move ecr gc build job to ecr gc workflow (#38523)

seemethere committed 4 years ago

8251f187

Fast tanh for the LLVM backend. (#39528)

zheng-xq committed 4 years ago

83dd5663

c10/util/complex_math.h and c10/util/complex_utils.h should not be individually included (#39276)

xuhdev committed 4 years ago

df2d1972

Cleanup rref_impl (#39530)

rohan-varma committed 4 years ago

397b24bb

[predictor] move fblearner/predictor to platform009

HC Zhu committed 4 years ago

a2125135

[ROCm] Enable AMP autocast tests on ROCm (#39616)

rohithkrn committed 4 years ago

ab6c447f

Revert D21930435: Revert D17923732: Optimize GroupNorm on CUDA

ezyang committed 4 years ago

cc2f7fa5

add dtype checks for scatter/gather family of functions. (#38646)

nikitaved committed 4 years ago

e4f9c74d

Remove duplicate 'with_gil' declaration. (#39540)

gchanan committed 4 years ago

e3e8f24c

Revert D17923732: Optimize GroupNorm on CUDA

ezyang committed 4 years ago

a83f7a1d

Add error message when negative stride is passed to as_strided (#39508)

wjgan7 committed 4 years ago

e41fe608

add overload name for min/max with list input (#39614)

linbinyu committed 4 years ago

820e81ba

[futures] Add c++ ivalue::Future collectAll() helper (#39119)

jjlilley committed 4 years ago

b83fed8d

Newer Older