strisunshinewentingwang/pytorch

Pull Requests Commits

Add `torch.flip{lr, ud}` (#38599)

kshitij12345 committed 4 years ago

97333909

[iOS] Disable depthwise3x3_winograd on iOS (#39591)

xta0 committed 4 years ago

4ec86ca5

Use atomic operations to manipulate current RPC agent (#39663)

lw committed 4 years ago

7d85e770

torch.multinomial : fast-path for replacement=False (#39636)

kshitij12345 committed 4 years ago

af05158c

Fix error handling for rpc.remote (#39605)

mrshenli committed 4 years ago

338a1ccc

Kill dead pairwise ops in THC (#39070)

ShawnZhong committed 4 years ago

aa5ccf96

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

Newer Older